kant and the meaning of existence: a modal account by ian...
TRANSCRIPT
KANT AND THE MEANING OF EXISTENCE: A MODAL ACCOUNT
by
Ian Blecher
, Yale University, 2000
, University of Pittsburgh, 2010
Submitted to the Graduate Faculty of
the Kenneth P. Dietrich School of Arts and Sciences
in partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
University of Pittsburgh
2012
ii
UNIVERSITY OF PITTSBURGH
KENNETH P. DIETRICH SCHOOL OF ARTS AND SCIENCES
This dissertation was presented
by
Ian Blecher
It was defended on
August 2, 2012
and approved by
John McDowell, Distinguished University Professor of Philosophy
Robert Brandom, Distinguished Professor of Philosophy
Nicholas Rescher, Distinguished University Professor of Philosophy
Marah Gubar, Associate Professor of English
Dissertation Director: Stephen Engstrom, Professor of Philosophy
iii
KANT AND THE MEANING OF EXISTENCE: A MODAL ACCOUNT
Ian Blecher, Ph.D.
University of Pittsburgh, 2012
It is a distinctive claim of Kant’s Critique of Pure Reason that what exists outside the
mind is always represented under a modality — i.e. as possible, actual or necessary.
The interest of this claim is not widely appreciated. Most commentators have
ignored it; a few have rejected it out of hand. Since the Critique presents modality
as a basic aspect of human knowledge, however, this is a serious oversight. My
dissertation is an attempt to rectify it. The main idea is that, for Kant, the
knowledge of what exists is connected with a certain kind of progress in the mind —
a progress from the capacity to know (possibility) to the act of this capacity,
(actuality), and finally to the perfection of that act (necessity). To the extent that the
representation of this progress figures in our knowledge of what exists, such
knowledge is thus at least implicitly modal. I argue, however, that Kant also
intends something stronger: viz. that knowledge of what exists is constituted by its
representation within the progress of modalities. It follows that modality is not just
one feature of this knowledge among others, but its characteristic form.
iv
TABLE OF CONTENTS
PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
Notes on sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
Notes on translation and terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x
INTRODUCTION: Existence as Modality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
PART 1: Modality and the Act of Judgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
CHAPTER 1: The Formal Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
CHAPTER 2: The Real Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
PART 2: Modality and the Doctrine of Judgment . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
CHAPTER 3: The Modal Schemata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
CHAPTER 4: The Principles of Modality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
APPENDIX: The Functions of Modality and Relation . . . . . . . . . . . . . . . . . . . . . . . .71
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
v
LIST OF TABLES
1. Categories of modality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2. The functions of modality and relation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
vi
PREFACE: ACKNOWLEDGMENTS
Mentor, save us from ruin!—Odyssey, .208.
1. It may seem presumptuous to begin with thanks, for the dissertation that follows
may not be very good. And even if it is not very bad, it is still only a dissertation:
essentially a school assignment written to satisfy a scholastic requirement. (The fact
that its author is 33 years old only makes things worse.)
The purpose of these acknowledgements is not, however, to divide the spoils of
a philosophical victory. It is to express gratitude for an education. My hope is that I
may someday prove worthy of it.
2. I want to begin by acknowledging a debt to my Committee. Robert Brandom,
John McDowell and Nicholas Rescher have been a great help to me, both in their
comments on my work, and in their courses and writings on Kant’s philosophy.
Marah Gubar has given much-needed encouragement and professional advice.
Stephen Engstrom directed the dissertation with extraordinary care; his influence
will be felt in whatever truth it contains.
3. I also want to thank many other friends, colleagues, teachers, administrative and
mental health professionals for aid and comfort, including — but not limited to —
Alp Aker, James Allen, Jacob Blecher, Joel Blecher, Jochen Bojanowski, Matt Boyle,
Chris Campbell, Jim Conant, Mike Caie, Brandon County, Damian Da Costa, Michael
Della Rocca, Cian Dorr, Amy Enrico, Sam Floyd, Anton Ford, Hannah Frank, Chris
Frey, Jennifer Frey, Jessica Gelber, Tamar Gendler, Susan Gillo, I-Huei Go, David
Goldsmith, Matthias Haase, Karsten Harries, Collie Henderson, Ulf Hlobil, David
Horst, Connie Hrabovsky, Dale Kiefer, Irad Kimhi, Andrea Kern, Christian
Preface vii
Kietzmann, Sandy Kornblith, Thomas Land, Ben Laurence, Doug Lavin, Hans
Lottenbach, Lissa Merritt, Jess Moss, Evgenia Mylonaki, Sasha Newton, Katharina
Nieswandt, Julie Oppenheimer, Kathy Rivet, Sebastian Rödl, Karl Schafer, Kieran
Setiya, Jamsheed Siyar, Daniel Smyth, Chandra Speeth, Judy Suh, Daniel Sutherland,
Rita Svetlova, Michael Thompson, Markos Valaris, Herb Wilson and Jessica Winter.
4. I owe a special debt to the students and faculty of the German Department at the
University of Pittsburgh. It is because of them that I have been able to read Kant in
the original, and to study his philosophy in Germany. Yet the future of the
Department’s graduate program is now in doubt; I entreat the Dean to reinstate it
as soon as possible.
5. I thank the University of Pittsburgh, the Woodrow Wilson Foundation, the
Deutscher Akademischer Austausch Dienst, and my family for crucial material
support.
6. I thank my parents for never asking when this would be finished.
7. I dedicate this dissertation to my wife, Sara Nichols — who does not care
whether it is good.
Telle petite action ingénieusement, exquisement bonne, qu’elle aaccomplie pour moi, telle géniale attention, tel geste familièrementsublime, révèlent une compréhension plus profonde de l’existence quetous les traités de philosophie.
Preface viii
NOTES ON SOURCES
1. References to the Critique of Pure Reason follow the pagination of the two original
editions, “A” (1781) and “B” (1787). Passages common to both editions are cited
accordingly. I occasionally refer to the 1929 translation of Norman Kemp Smith
(London: Macmillan) by the abbreviation .
References to Kant’s other works include the title (or its abbreviation), the
volume of the “Academy” edition () in which it appears, and the relevant page
numbers in that volume.
The following abbreviations are used throughout. I note original dates and places
of publication (where applicable), together with the relevant volume of the
Academy Edition.
= Akademieausgabe von Immanuel Kants gesammelten Werken in 29volumes. Berlin: various, 1900-.
= “Der einzig mögliche Beweisgrund zu einer Demonstration desDaseins Gottes” [“The Only Possible Ground of Proof for aDemonstration of the Existence of God”]. Königsberg, 1763. 02.
Br = Briefwechsel [Correspondence]. 10.G = Grundlegung zur Metaphysik der Sitten [Groundwork of the Metaphysics
of Morals] (second edition). Riga, 1786; 04.KpV = Kritik der praktischen Vernunft [Critique of Practical Reason]. Riga, 1788.
05.Log = Immanuel Kants Logik, ein Handbuch zu Vorlesungen [Logic], edited by
Gottlob Benjamin Jäsche. Königsberg, 1800. 09. = Kritik der Urteilskraft [Critique of Judgment] (second edition). Berlin
and Libau, 1793. 05. = Die Metaphysik der Sitten in zwei Teilen [The Metaphysics of Morals]
(second edition). Königsberg, 1798. 06. = Opus postumum. 1790-1801. 21 and 22.Prol = Prolegomena zu einer jeden künftigen Metaphysic, die als Wissenschaft
wird auftreten können [Prolegomena to any Future Metaphysics That WillBe Able to Come Forward as a Science.] Riga, 1783. 04.
Preface ix
R = Reflexionen Kants zur kritischen Philosophie. [Kant’s Reflexions on theCritical Philosophy]. 14-19. Citations include dates given byAdickes’s Schriftphasen. 14: -.
= “Uber den Gemeinspruch: Das mag in der Theorie richtig sein, taugtaber nicht für die Praxis”. [“On the Common Saying: That May BeCorrect in Theory, but It Is of No Use in Practice”.] Berlin, 1793. 08.
2. The Blomberg, Philippi and Vienna Logics are based on transcripts of lectures
delivered in the 1770s and 1780s (see 24). Their fidelity may well be doubted,
but I have referred to them mainly where they lend support to views Kant also
expresses in his published writings.
3. References to works of medieval and early modern philosophy are given in
footnotes, usually with the relevant section number.
4. References to Plato follow the traditional (“Stephanus”) pagination. References to
Aristotle include the traditional titles, divisions and “Bekker” lines.
5. References to contemporary works are given in the text by author, original year
of publication and original page number. A more complete bibliography follows.
6. I am grateful to the Web site Kant in the Classroom for valuable bibliographic
information.
Preface x
NOTES ON TRANSLATION AND TERMINOLOGY
1. All the translations that appear in this dissertation are my own, though in some
cases I have consulted English versions. (I have relied especially on .)
2. Some especially difficult or important words or phrases are accompanied by the
original German, Latin or Greek in brackets. I have modernized spellings in
accordance with the neue Rechtschreibung.
3. Kant and his contemporaries used two forms of typographical emphasis:
Sperrdruck, represented in by increased space between letters, and Fettdruck,
represented in with boldface. For simplicity’s sake, I have not preserved this
distinction: all emphasis has been put in italics (including my own). I also use
italics to mark certain concepts (such as existence) and names (as in “the word
modality”).
4. I have translated Kant’s word Erkenntnis as knowledge. This is unsatisfactory for
at least two reasons. First, the English word knowledge suggests a state of mind, or a
disposition, while for Kant Erkenntnis is almost always active; an Erkenntnis is an
act of knowledge (I myself will make much of this). Second, Kant indicates that an
Erkenntnis can be false (see A58/B83), while knowledge is supposed to be factive.
For these and other reasons recent translators have preferred to render Erkenntnis
as cognition. That is understandable, but I am still not persuaded. Cognition is not a
term of ordinary English; even to specialists it suggests something strange and
technical. But for Kant Erkenntnis is supposed to name an ordinary concept, a part
of the epistemology of everyday life. And the closest thing we have to it is
knowledge.
5. In Chapter One, I make use of the Greek terms dunamis and energeia; as Frede
(1994) has noted these are very difficult to bring into modern English. The former
Preface xi
might, for example, be called a potentiality, the latter an actuality. I have translated
energeia as act, however, to distinguish it from a mere occurrence; to be consistent
with this, I have translated dunamis as capacity.
6. Kant himself uses a number of words that might sometimes be translated as
capacity — Fähigkeit, Vermögen, Kraft, Funktion, Befugnis, Kunst and so on. I have
tried to keep these separate as far as possible; in particular, I always translate
Vermögen as faculty, Kraft as power and Funktion as function. It is worth remembering,
however, that Kant’s own usage varies; there is no reason to assume that he always
intended Vermögen (for example) in the same sense.
7. He will occasionally shift between Latinate and Germanic synonyms — for
example, between Relation and Verhältnis, or between Position and Setzung. It is
sometimes claimed that these shifts are significant, and I have noted them where
appropriate. That is not to say, however, that every shift in terminology is
significant. Kant’s usage varies for all sorts of reasons, only some of which have to
do with philosophy.
INTRODUCTION: EXISTENCE AS MODALITY
The modality of judgments is a most peculiar function. . .—A74/B99.
0.1
1. Although this dissertation is broadly concerned with the meaning of existence in
Kant’s theoretical philosophy, the concept is treated throughout from the point of
view of a particular theory of modality. This theory will be characterized in a
number of ways in what follows, but its basic claim is that judgments are modal as
such. By way of an introduction, I will say a little about what that comes to; first,
though, some historical context.
2. The noun modality was unknown before the modern period.1 It is possible that
Kant invented it, though he never claims credit for having done so.2 In any case, it
did not name a standard topic in logic or epistemology before his time. (The
concepts that today are called modals or modalities — for example, possibility and
necessity — were known more generally as modes, i.e. determinations of attributes.3)
By contrast, the adjective modal had been used since late antiquity. Its original
purpose had been to mark off a special branch of Aristotle’s logic, the so-called1This has been much obscured by a tendency to speak blithely of, e.g. Aristotle’s theory of modality,
or Scholastic theories of modality. It may be that a theory of what we call modality can be found inAristotle, or among the Scholastics, but it should not be assumed that it was, for them, a theory of thiskind.2Here I am adapting a suggestion of Rainer Specht. (“Modalität”, Historisches Wörterbuch der
Philosophie 6.) Specht himself cites .. Krug (Allgemeines Handwörterbuch der philosophischen Wis-senschaften. Leipzig, 1838, 5: 43). He does not note — but it may be significant — that the wordModalität does not appear in Kant’s own writings until the 1770s, when he was writing the Critique ofPure Reason.3For a classical definition, see Descartes, Principles of Philosophy sec. 56
Introduction 2
modal syllogistic;4 by the modern period, though, it had come to describe more
broadly any judgment in which the relation between the predicate and the subject
concept was itself represented under a determinate concept — paradigmatically that
of possibility or necessity, but in principle almost anything.5 This was opposed to
non-modal judgment, in which the relation was represented simpliciter. (Thus
Gesner: “a judgment [enunciatio] is modal if, e.g. I do not say ‘Offenses come’ but
rather ‘It must needs be that offenses come’”.6)
3. The relations between these two species of judgment were seen to raise a number
of complications, and controversies about them have proliferated since antiquity7 —
but these may be set aside here. The important thing is that, within the tradition, a
judgment was not assumed to involve modal representations. (Indeed, non-modal
judgment was considered more basic, because it was in one sense less qualified.)
All this may seem perfectly natural; certainly it is not incoherent. But it is not
Kant’s theory of modality. Kant’s theory, as I indicated, is that all judgments are
modal — they all have modality.
4. The theory can be taken in a strictly logical sense, as a theory about the kinds of
concepts that are necessary for a judgment to constitute a complete thought, and to
behave properly in the accepted forms of inference. In this sense, the theory says
that a judgment is well-formed just in case the predicate is related to the subject
concept under a further concept, the modality. (In fact, the theory goes further than
this, for it also specifies which concepts can serve as modalities,8 but that is not4Aristotle did not use this, or any other word for it. But the Prior Analytics, among other works,
contains an extended discussion of the behavior of necessary and possible premisses in the forms ofinference enumerated in the Organon (see especially A.9 ff).5Crusius includes essentiality and naturalness among these concepts (Weg zur Gewissheit und Zuverläs-
sigkeit, sec. 228). Vollgnadt includes any adverb — for example, wisely (Decas disputationum logicarumpublico Doctorum examini in celeberrima Rosarum Academia proposita. . . 6).6Primae lineae isagoges in eruditionem universalem 2, sec. 1051. Kant had studied this work; see
V-Lo/Blomberg, 24: 74.7The most notorious question concerns the “Two Barbaras” of the Prior Analytics, A.9. For an
illuminating discussion, see Rescher (1964).8See, e.g. A80/B106.
Introduction 3
important here.)
This theory also is not incoherent. It is, however, revisionary. For example, it
will posit a tacit modal representation (such as actuality9) in the judgment that
Caius is a man. Of course, not all revisionary theories are implausible.10 But there
is a presumption against them. Other things equal, a non-revisionary theory is
always preferable. The problem for any revisionary theory is thus to explain why
other things are not equal. Unfortunately, Kant does not do this — at least not
directly. (He does not even acknowledge that there is a problem.) But I think it is
possible to construct an explanation on his behalf, from materials provided in the
Critique of Pure Reason and other, related texts.
5. A crucial step is to see that the problem remains intractable so long as Kant’s
theory of modality is taken in a strictly logical sense. I do not mean that it should
never be taken that way. There are places in the Critique, and in the related texts,
that seem to require it; and even if there were not, it would sometimes be useful for
us to focus on the significance the theory has for logic. Rather, I mean that the
theory can be taken in a broader sense; and when it is taken in this sense the
problem becomes more tractable. (The logical sense will then be seen as a special
application of the broader theory.)
The broader sense cannot be explained in any detail here. It is the subject of
the entire dissertation. But the main point can be put like this:
Kant says that theoretical knowledge is knowledge of what exists outside the mind,
and ultimately in space and time.11 If (as he also says) this knowledge consists in9Which modal representation is posited often depends on the judgment’s context. But this com-
plication can be set aside for the moment.10For example, a theory of the quantity of knowledge is not implausible just because it interpretsthe statement “Bodies are divisible” as the judgment that all bodies are divisible; in fact this is thenatural interpretation.11The contrast is with practical knowledge, which is knowledge of what should exist, though it may
not (A633/B661). While the Critique of Pure Reason has something to say about such knowledge, it isnot the central topic.
Introduction 4
judgments of a certain kind, then such judgments will as such involve a
representation of the existence of their objects. And the point will be that this
representation is as such modal.
In other words, theoretical judgments have modality because existence itself is a
modal concept — like possibility or necessity. Moreover, it bears a systematic relation
to those concepts, so that an understanding of the meaning of existence will depend
on — and at the same time enable — an understanding of them, and of modality in
general.
Why Kant thinks these things is a question I will try to answer in the body of
the dissertation.
0.2
1. It is divided into two parts, broadly corresponding to Kant’s own procedure in
the central chapters of the Critique of Pure Reason, the “Transcendental Analytic”.
a. In Part One, I consider two respects in which an act of judgment is said to have
modality: (1) with respect to its logical form; and (2) with respect to its content. In
Chapter One, I isolate a merely formal concept of modality, and I argue that for
Kant this concept is connected with the logical representation of a certain kind of
progress in understanding. In Chapter Two, the argument is expanded through
consideration of the modal content of judgments, the categories of modality. In
particular, I argue that these categories are concepts of progress in the representation
of existence (in one of its senses) in an understanding like ours: namely, one that is
self-conscious in a certain sense, and also finite, i.e. dependent upon sensible
representations.
b. In Part Two, I apply this general account of the modality of judgments to the
sensible conditions under which human judgment occurs — and especially the most
Introduction 5
general of these, viz. time. In Chapter Three, I discuss the capacities that make
possible the representation of the categories of modality in relation to time, the
schemata of modality. I argue that — despite appearances — these capacities add
nothing to the categories themselves, but only provide a necessary condition of their
use. In Chapter Four, this use is presented in three principles, the “Postulates of
Empirical Thinking in General”. These principles are widely regarded as trivial (if
not false); I argue, however, that they have been misunderstood. Ultimately, their
significance lies not in representing the concepts of the possibility and actuality and
necessity of existence in time, but in realizing those concepts as modalities of human
knowledge.
2. My treatment of these topics is far from comprehensive. Many important
questions are omitted, and others are addressed incompletely. (To take just one
example, I say nothing about the “Refutation of Idealism”, though it may seem to
be a natural topic for me.) Partly this is because I have written a dissertation and
not a book; I have been constrained throughout by statutory deadlines and
inexperience in addition to my more particular shortcomings. Partly, though, it is in
the nature of the subject. It is clear from the beginning that the question of the
meaning of existence, and Kant’s peculiar analysis of it, can be taken in many ways.
My hope is that, in offering one approach, I have not foreclosed others.
PART ONE: MODALITY AND THE ACT OF JUDGMENT
CHAPTER ONE: THE FORMAL MODALITIES
One who is learning a science has the capacity to know it in a differentsense than one who — while already possessing the knowledge — is notactually exercising it.
—Aristotle, Physics 8.4, 255a33 f.
1.1
1. It is a distinctive feature of Kant’s theory of logic that all judgments, as such,
have a modality.
In other words, modality belongs to what is called the logical form of judgments
— the manner in which representations are ordered in them. This is different from
the modality of a judgment’s content. Kant says that judgments can include the
concept of an object as possible or impossible; existent or non-existent; or necessary
or contingent. Obviously these are modal concepts; he calls them the categories of
modality. As contents of judgment, however, they are not part of his theory of logic,1
and I will not consider them in this chapter. I will consider them in Chapter Two.
2. Formal modality determines what Kant calls an affirmation or a negation, and
what I will call more generally a thought. This thought can be a simple predication
(as in “All men are mortal”), but it can also be complex (“If all men are mortal,1See A54/B78: “[G]eneral logic abstracts from all content. . . ” The distinction between modality of
content and modality of form has sometimes been interpreted as a distinction between modality de reand modality de dicto. It is worth noting, however, that nothing in the Critique of Pure Reason requiresthat interpretation, and in fact it is inept to the extent that modality de dicto pertains to the contentof judgment. A closer comparison with logical modality would be Frege’s concept of sentential force.But this is inexact, since for Frege force does not have modal significance.
1. The Formal Modalities 8
then Caius is mortal”; “Either all men are mortal, or all men are non-mortal”).2 A
formal modality is the representation of the value of this thought — that is, its
“value in relation to thinking in general” (A74/B100).
Although he does not say this explicitly, Kant’s view seems to be that
judgments differ from mere thoughts by their inclusion of these representations.
For in a judgment a thought is not just entertained, but assigned a certain kind of
significance. This significance is subjective, in the sense that it pertains to the
subject’s own conception of his thought, and not to the thought itself. But — as I
will argue — it is not merely psychological; it is not just a private opinion. It is,
paradigmatically, a recognition of the judgment’s role in human knowledge.
3. Kant distinguishes three special formal modalities, corresponding to the three
pairs of modal categories I mentioned in 1.1.1. A judgment is problematic, he says, if
it represents a thought as merely possible (or arbitrary); it is assertoric if it
represents the thought as actual (or true);3 and it is apodeictic if it represents the
thought as necessary (A74 f./B100 f.).4
Since Kant’s view is that all judgments are modally determinate with respect to
their logical form, the proposition I am going to consider is that
(M) All judgments are as such problematic, assertoric,or apodeictic.
Although his descriptions of these modalities suggests that he takes this disjunction
in an exclusive sense — at least as regards problematic and assertoric judgments —2Today we would call these thoughts propositions. For Kant, however, a proposition (Satz) is an
assertoric judgment rather than mere relations of concepts (Log sec. 30: n2, 09: 109; cf. R 3111;V-Lo/Wiener, 24: 934).3Note that for Kant, an assertoric judgment is not an assertion in the contemporary sense of
“say[ing] something in such a manner as deliberately to convey the impression of saying it with theoverriding intention of saying something true,” as Dummett defines it (1981), 300. Nor is it this actin foro interno. It is simply the representation of a thought as true.4In the Logic manuscript prepared by .. Jäsche, Kant compares these values to those associated
with opinion, belief, and science [Wissen] respectively (Log sec. , 09: 65 ff.). The comparisoncan be instructive, but it is important to remember that while the latter have primarily a subjectivesignificance, the former have primarily a logical one: the fact that, for example, problematic judgmentsexpress opinions is not part of the account of them as judgments.
1. The Formal Modalities 9
I will consider an alternative interpretation below (sec. 1.3.1).
4. I suggested that M is distinctive in the history of logic. In fact it is distinctive in
two respects. It is distinctive in its terminology: though they have cognates in the
tradition,5 the terms problematic, assertoric, and apodeictic are Kant’s invention.6 And,
as this suggests, M is also distinctive in its conception of judgment.
The easiest way to bring this out is to consider a traditional conception of
formal modality:
The representation of the manner in which the predicate does or doesnot belong to the subject is the determination of the concept of theircombination and its negation (modus formalis). Either a judgment has such adetermination or it does not. The former is an impure judgment (iudiciummodale, modificatum, complexum qua copulam), e.g. ‘This world necessarilyexists’, ‘It does not necessarily exist’. The latter however is a purejudgment (iudicium purum).
This is from Georg Friedriech Meier’s 1752 Auszug aus der Vernunftlehre, an
influential textbook of the period.7 The most obvious difference between Kant and
Meier is that Kant does not recognize a class of non-modal judgments — indeed,
Kant himself remarks, in a handwritten note on this passage, that “Without
modality, no judgment at all would be possible” (R 3111, 16: 663, 1770s).8
Here, however, I want to focus on a subtler difference, a difference in their
conceptions of logical form. Kant points to this himself, in a lecture apparently on5Aristotle, for example, speaks in the Topics of problemata (101b29), i.e. what is to be vindicated or
refuted in a dialectical argument; and, in the Posterior Analytics, of apodeixis (Α.2, 71b9 ff.), i.e. what isdemonstrated by means of a proper syllogism.6See Tonelli (1966), 156.7Sec. 309. The emphases are Meier’s.8See the Introduction, 0.1.3, above.
1. The Formal Modalities 10
Meier9 from about 1780. Logicians in the tradition, he says,10
did not take the division of that which concerns the modality ofjudgments so precisely as we do, but called every concept of combination‘modality’. [Take] for example [the judgment], ‘the world exists in anecessary manner’. Here the word in a necessary manner was themodality. But can logic really judge whether a thing is necessary or not?No, for it has nothing to do with things and their necessity. Accordinglyit can ask only whether a judgment is expressed with necessity or not.[V-Lo/Wiener, 24: 935, emphasis added.]
In denying that the representation of necessary existence belongs to logic, Kant
is not denying that it is a modality in any sense.11 He is denying that it is a formal
modality — that it belongs to the logical form of judgments (how they are
expressed). For, he thinks, the representation of the manner in which a thing exists
pertains to the content of judgment, and not the mere form of representation.
By itself, this would not be noteworthy. Apparently Meier too had denied that
representations of possible and necessary existence belong to logic in the strictest
sense, for he had classified the judgments that contain them as impure.12 And this
was a widespread view in the 18th century. According to .. Lambert, for example,
modal concepts “belong to ontology, and do not depend merely on the external [sc.9He does not explicitly mention Meier in the lecture, but it is clear from the example he gives that
he has him in mind. The Auszug aus der Vernunftlehre had formed the basis of Kant’s standard courseof lectures on logic since the 1750s (see : ), and he knew it extremely well: by the end ofhis teaching career his own copy was so heavily annotated that an acquaintance, Count Purgstall, wasmoved to describe it in a letter to .. Kalmann:
Kant reads from an old Logic — of Meier’s, if I am not mistaken. He always has thebook with him. It appears so old and dirtied, I believe he has brought it to class dailyfor 40 years now; all the leaves are filled with small writing in his hand, and, what’smore, notes are pasted to many of the printed pages, and many of the original lines arecrossed out, so that, as this suggests, almost nothing is left of Meier’s work. [April 30,1795. Cited in Malter (1990), 421.]
These annotations would form the basis for the 16th volume of the Prussian Academy’s edition ofthe works of Kant.10Transcripts of Kant’s lectures are not necessarily accurate; in this case, however, the point is alsoattested in his published works.11Sellars seems to read him that way (1968), 54 f.12Meier never defines pure and impure in the Auszug aus der Vernunftlehre. But the glosses he offers— e.g. modificatum — suggest he thinks of these concepts as derivative.
1. The Formal Modalities 11
logical] form of knowledge”.13 And according to Baumgarten, “modes [sc.
modalities] are treated contingently in logic” — so that logicians may omit them if
they like.14 Wolff did omit them.15 Nor was he alone in this.16
What is distinctive in Kant’s logic is the positive characterization of the logical
form of judgments as involving modality. Traditionally, logicians had characterized
judging, as opposed to mere predicating, in terms of what they called assent or
positing (Latin: assensus, ponere aliquid; German: annehmen).17 This shares something
with Kant’s characterization: to assent is to represent a thought as true rather than
arbitrary (cf. R 2506, 16: 397, where an analogy is explicitly drawn). But there
is also an important difference: assent was not a proper modality. It did not
represent a thought as actual; it was not opposed to the representation of the
thought as merely possible. It was opposed, rather, to the withholding of assent, or
the suspension of judgment.
There is nothing obviously wrong with the traditional characterization. Kant
himself had accepted it for much of his career.18 So why did he think it was
unsuitable to the logical theory of the Critique of Pure Reason? And why did he
introduce the formal modalities in its place? In what follows, I will try to answer
these questions by considering the role of logic in Kant’s broader theory of
knowledge.13Neues Organon oder Gedanken über die Erforschung des Wahren und der Unterscheidung von Irrtumund Schein , sec. 137.14Acroasis Logica in Christianum L. B. de Wolff, sec. 160.15They are, however, discussed at length in his Ontologia, as Longuenesse notes (1998), 158n.16See Adickes (1887), 38.17See Aristotle, De Interpretatione, 17a25 ff.; Locke, An Essay Concerning Human Understanding .,
sec. 3 et passim; Hume, A Treatise of Human Nature, ., sec. 7; Wolff, Logica, sec. 1009; and Meier,Auszug aus der Vernunftlehre, sec. 168. This is not to say a distinction between judgments and thoughtswas rigorously maintained; logic textbooks often referred to what I am calling thoughts as judgments.Nonetheless, it was maintained, contrary to what is sometimes thought. For a discussion of the point,see Owen (2007), 412 ff.18See the Blomberg transcript of his logic lectures from 1770 ( 24: 277) and the Philippi transcriptof lectures from the following year (cf. 24: 463). It was not until the Vienna lectures of 1780 —from which I quoted — that he began to speak of modality in the logico-formal sense; and it was onlyin 1781, with the first edition of the Critique of Pure Reason, that he introduced the terms problematic,assertoric and apodeictic to designate formal modalities.
1. The Formal Modalities 12
1.2
1. The natural place to begin is section nine of the Critique, where the formal
modalities are introduced.
The purpose of this section is to provide a systematic representation of the
logical form of understanding or judgment.19 This is accomplished by means of a
table comprising twelve moments (or specific forms20) of judgment, three each
under four headings: quantity, quality, relation and modality. The Table appears in
both editions without a title; for convenience I will refer to it as the Table of Logical
Forms.21
It is accompanied by four short appendices. On the surface, these seem to
consist in small correctives to traditional logical doctrine, for example, that singular
judgments (e.g. “This man is mortal”) are formally different from universal ones
(“Every man is mortal”); and that affirmative judgments (e.g. “Caius is mortal”) are
formally different from infinite ones (“God is non-mortal”);22 and Kant probably
does take these points to have intrinsic significance for logic. But he intends
something more too.
Traditionally, Kant says, a judgment’s logical form was conceived as the19For Kant the faculty of understanding and the faculty of judgment are the same. To understand,for him, means to know by judging. See, e.g. A69/B94, A81/B106, A97.20Because all judgments originate in understanding, they share a generic form. This might be calledthe form of judgments. It is natural, however, to speak of the moments in which the form of judgmentsconsists as themselves forms — for example, the assertoric form of judgment. This way of speaking willnot cause confusion so long as it is not taken to imply an irreducible plurality of such forms.21It is called the Logical Table of Judgments in the Prolegomena (sec. 21, 04: 302). Perhaps for thisreason commentators have tended to call it the Table of Judgments. That title, however, is imprecise,since judgments have both form and content. Probably it ought to be called the Table of Moments ofthe Logical Function of Understanding in Judgments (cf. A69/B94), but that would be more precisionthan is necessary here.22Infinite judgments can also be distinguished from merely negative ones. The former express anaffirmative representation of a negated predicate, e.g. that non-mortality belongs to God. By contrast,the latter (e.g. “God is not mortal”) express a negative predication, i.e., that mortality does not belong toGod. It is logically possible, then, for negative judgments to be infinite, e.g. “Caius is not non-mortal”.Perhaps the reason Kant does not draw attention to this contrast is that it can be drawn in terms ofthe relations between judgments; whereas the contrast between affirmative and infinite judgments canonly be drawn by appeal to the relation of judgments to objects.
1. The Formal Modalities 13
manner in which it entered into the commonly accepted forms of inference — e.g.
Barbara or Darii. Judgments that entered into the same forms of inference in the
same manner were assigned the same logical form; the totality of such forms was
given by the totality of forms of inference.23 Although Kant regards this conception
as suitable to the narrow purposes of logicians, he insists that theoretical philosophy
must have a richer conception of logical form: viz. as a formal specification of the
faculty of knowledge itself. The forms of judgment, in other words, have to be
conceived as forms of knowledge, to which certain contents will be peculiarly
suited. (Thus he will speak, for example, of regarding judgments — for purposes of
the Table — “not merely according to inner validity, but also as knowledge in
general” [A71/B96].) It is this conception that is the source of the Table’s distinctive
features, including — I will argue — the theory of formal modality.
Take the example of the affirmative and infinite forms of judgment, which Kant
distinguishes with respect to the quality of judgments. In traditional logic, he says,
these were treated as equivalent, since they enter into inference forms in the same
way. It makes no difference, with respect to inferences, whether the predicate
concept is positive or negative in itself. But it does make a difference to the
knowledge the judgments represent. It is not just that someone will typically assign
differential value to positive predicates (in the sense that he will find it more
interesting to know that something is P than that something is non-P); the
knowledge itself will have a different formal character. According to Kant, affirmative
judgments represent knowledge of a positive determination of the subject concept;
by contrast, infinite judgments are “limitative only” (A73/B98) — they merely
exclude a particular determination from the predicate through which the object is23This characterization of traditional logic is not completely accurate. Early modern logic textbookstypically consisted in a hodgepodge of commonly recognized forms of judgment described against thebackground of a vaguely systematic account of the roles of some of them in commonly recognizedinference forms. Perhaps Kant’s point is not that logicians were ignorant of forms of judgment notconnected with syllogistic, but that they could not explain why these forms had to be included in thetheory of logic. (This, at any rate, seems closer to the truth.)
1. The Formal Modalities 14
represented. (Consequently, Kant associates the former with the category of reality,
the latter with the category of limitation.)
2. It can seem as though these considerations do not bear on the formal modalities,
for judgments that differ in formal modality do differ in their inferential behavior.
Thus Kant will say that “both judgments, the relation of which constitutes the
hypothetical judgment. . . , and likewise in whose reciprocity the disjunctive
judgment consists. . . , are in their entirety only problematic”; also: “in a
hypothetical syllogism, the antecedent occurs in the major premiss problematically,
in the minor assertorically” (A75 f./B100 f.). (He does not mention apodeictic
judgments in this connection, but presumably he takes them to correspond to the
conclusions of categorical syllogisms.) For this reason, Longuenesse, among others,
has argued that “the modality of a judgment is determined by its relation to the
forms of thought involved in deductive reasoning (judgments and syllogisms).”24
Clearly, there is a sense in which she is correct. The modality of a judgment can
be taken to correspond to the manner in which it enters into forms of inference. It
is important, however, not to overstate the significance of this. For it is not a
sufficient explanation of M — it does not show why the “relation to the forms of
thought involved in deductive reasoning” must be conceived modally, rather than
merely in terms of giving and withholding assent. For example, there does not
seem to be any prima facie reason that the relation of categorical judgments to the
form of hypothetical syllogisms could not be specified by observing that assent is
withheld from the antecedent in the major premiss, and given in the minor.25 (This
is how it typically was specified in traditional logic.) To explain the relation of
modalities to the forms of inference there must first be an account of the modality24(1998), 159; cf. Longuenesse (2005), 99 and 190 f.; Vuillemin (1960-1), 322 f.25One answer would be: Kant already construes “the forms of thought involved in deductivereasoning” in modal terms. As a matter of fact I think he does construe them this way (perhaps thatis what Longuenesse has in mind; it is, at any rate, a point that has been suggested by some Kantinterpreters. But Kant cannot take this for granted: the construction is not self-evident; nor does ithave a basis in any system of logic prior to the Critique.
1. The Formal Modalities 15
of judgments in general.
1.3
Kant gives such an account. The main feature of it is described toward the end of
section nine:
[E]verything. . . is incorporated into the understanding stepwise, so thatsomething is first judged problematically, then perhaps is also acceptedassertorically as true, and finally is asserted as inseparably bound upwith understanding, i.e., as necessary and apodeictic. . . [A76/B101.]
— and, he says, it is this that explains why the formal modalities have to be
included in the Table of Logical Forms. (His exact words are: “Since now
everything here is incorporated into the understanding stepwise. . . these three
functions of modality can therefore be called so many moments of thinking in
general” [emphasis added].26) It is the only explanation he gives for their inclusion.
The passage is compressed, but the point seems to be that the formal
modalities, taken together, constitute a kind of progress in understanding; the
modality of a particular judgment would correspond to its place in this progress.
Moreover, in view of the at least implicitly self-conscious character of judgment, as
Kant conceives of it (B131 et passim), the modalities can be taken to represent this
progress, at least implicitly.
I want to draw two implications from this.
a. First, the modal values persist from one stage in the progress to the next. The same
possibility that is represented in a problematic judgment is also represented in an
assertoric judgment; the same actuality that is represented in an assertoric judgment26Some commentators have thought that Kant means to say something special about the formalmodalities in calling them “moments of thinking in general”. That is wrong. Many other passages(e.g. A78/B93) suggest that he regards all the forms of judgments as “moments of thinking in general”.Indeed, he appears to think that a moment of thinking in general just is a form of judgment.
1. The Formal Modalities 16
is also represented in an apodeictic judgment. Actuality includes possibility;
necessity includes actuality.27 The progress of modalities is not, therefore, a change
of mind. It is not like going from withholding assent to giving it.28
b. Second, the lower values anticipate the higher. The possibility that is represented in
a problematic judgment is not merely contrasted with actuality and necessity — as
affirmation, for example, is contrasted with negation. The problematic judgment
already includes the representation that it is the first stage in a progress that will
culminate in apodeictic judgment. The understanding does not just ‘fetch up’ at
apodeictic judgment, as if by accident. Apodeicticity is in view from the beginning,
as the constitutive purpose of all judgment.
In what follows, I will register these points by conceiving of the modalities in terms
of the Aristotelian distinction between capacities and acts.29 A problematic judgment
will thus be a capacity for the corresponding assertoric judgment; an assertoric
judgment will be a capacity for the corresponding apodeictic judgment.
Kant himself does not put things quite this way, and I could be accused of
imposing a certain metaphysic on the text. As a general point, though, he conceives
of understanding as a faculty (Vermögen) whose act is judgment. What I am
proposing is that particular acts of judgment can themselves be divided into
capacities and acts. (This will not make for profligacy in the account so long as the27Consequently, the disjunctions in M can also be taken inclusively (see 1.1.1 above). This is not anambiguity in the account. It is a question of emphasis. If the formal modalities are construed as formalclassifications of judgments, it will be natural to emphasize their distinctness. It would be confusing— not to mention tedious — if a single judgment had to be classified as problematic and assertoricand apodeictic. If, however, the formal modalities are taken as moments of the logical form of judgments,then it will be natural to emphasize their progressive unity — to speak of a problematic moment withinassertoric judgment, and an assertoric moment within apodeictic judgment.28It is natural to think that, when assent is given after being long withheld, this too is a kind ofprogress, for something must have been learned to justify a change of mind. This may be right. Thepoint, however, is that assent does not represent progress. The fact that it can be taken as an indicationof progress depends on the psychological history of the judgment in question, among other things.29What is the alternative? Possibility and necessity, at least, might be conceived as representing mereconceptual properties, like non-contradictoriness, or having-a-contradiction-as-a-negation. (This wasa traditional view.) The difficulty with such a conception — at least from Kant’s perspective — isthat it is unclear how progress could, in general, be effected from the former to the latter.
1. The Formal Modalities 17
further capacities are not construed as distinct agents, in Locke’s word, each of
which would have a distinct explanatory role in knowledge.30)
2. Before I say more, I want to head off a misunderstanding.
Progress in understanding might suggest a process of learning. The problematic
judgment would be a sort of hypothesis; the assertoric judgment would add to this
a determinate truth value (true, false); and the apodeictic judgment would represent
the assertion under an appropriate explanatory principle.
This certainly fits a common interpretation. Here is Jill Vance Buroker:
In problematic judgments one thinks or apprehends the judgmentwithout making a commitment to a truth value. . . Both assertoric andapodictic judgments involve assertions. . . 31
And this is Henry Allison:
problematic [judgments] are deemed capable of a truth value, though, asproblematic, this value is undetermined. By contrast, the truth value of[assertoric judgments] is determined. . . 32
It is true, of course, that human subjects learn — indeed they learn everything
they know. (Kant rejects innatism at B166 f.) But this does not make it logically
necessary to proceed stepwise through the modalities. It does not seem to be a
logical requirement that, before someone can learn whether something is true, he
must first represent it as a mere possibility.
And this has tended to make Kant’s theory of formal modality look like a kind
of psychologism. Here is how Norman Kemp Smith puts it:30This is from the Essay:For it being asked, what it was that digested the Meat in our Stomachs? It was a ready,and very satisfactory Answer, to say, That it was the digestive Faculty. What was it thatmade any thing come out of the Body? The expulsive Faculty. What moved? The MotiveFaculty: And so in the Mind, the intellectual Faculty, or the Understanding, understood. . .[., sec. 20.]
31(2006), 91.32(2004), 145 f.
1. The Formal Modalities 18
Kant’s remark [concerning the progress of modalities] is irrelevant andmisleading. The advance from consciousness of the problematic, throughdetermination of it as actual to its explanation as necessary, representsonly a psychological order in the mind of the individual.33
And this is Kneale and Kneale:
[Kant] says that the three ‘functions of modality’ which he distinguishesrepresent three ‘moments of thought’ which form a series. If this isindeed his doctrine, it is a bad instance of the corruption of logic bypsychology and epistemology.34
It would indeed be bad. It would be bad even as psychologism: for there is
nothing in empirical psychology to suggest that human beings typically proceed by
judging, e.g. “[The thought is merely possible:] this S is P”. . . “[The thought is actual:]
this S is P”. . . “[The thought is necessary:] Every S is P”. Many things are just
obvious; there is no need to approach them as a scientist approaches an
experimental question. In any case, I think another interpretation of A76/B101 is
possible — one that takes account of the tendency of human subjects to learn, but
that also construes the progress of modalities as an aspect of logical form rather
than a psychological tendency.
1.4
1. In Metaphysics Θ, Aristotle distinguishes two senses in which a capacity is said to
act. In a first sense, it involves what he calls a motion.35 His example is the process
of losing weight, i.e. an act of the body’s capacity for alteration under certain
conditions (e.g. diet and exercise) (Θ .6, 1048b19 ff.).36 The word alteration itself33(1918), 194.34(1985), 356.35Kinesis. Not all kineseis are physical motions; he will use the term also for what might be calledpsychological motions. Cf. Physics 3. For discussions, see Frede (1994); Kosman (1994).36Recent philological research suggests that this passage did not originally belong to the manuscriptfor Metaphysics Θ, and perhaps not to the Metaphysics at all. It is still generally agreed, though, thatthe passage was written by Aristotle, and it is clearly connected with themes that run through his
1. The Formal Modalities 19
suggests this: it is not a matter of a fatter thing’s going out of existence and a
thinner thing’s taking its place; it is a matter of a thing that was fatter now becoming
thinner, a transition between opposing states of a single thing.37
But — though Aristotle notes that “it is [widely] believed that an act is motion
most of all” [Θ.3, 1047a32 f.]) — he does not think that motion is an act in the
primary sense. In the primary sense, an act is not motion at all, for it involves no
alteration. “For example, at the same time we are seeing and have seen. . . are happy
and have been happy. . . are living and have lived” (Θ.6, 1048b22 ff., emphasis
added). These examples are not univocal, but the point seems to be that an act in
the primary sense is distinguished by being its own end (so that the act of seeing is
at the same time for the sake of seeing). By contrast, motions always pertain to some
further end, the achievement of which will bring the process to a natural close (so
that the weight loss will end when the excess weight is lost). To put this in
ontological terms: if a motion is a process by which a thing is altered, an act in the
primary sense consists in the capacity’s simply being what it is. (In some cases it
will be natural to say that it consists in the capacity’s being more fully what it is.) It
will be convenient to refer to acts in the primary sense as pure acts, and to use the
term act (simpliciter) for the genus governing both motions and pure acts. The
capacity for a pure act I will call a pure capacity.
There are many questions about Θ.6 that I cannot address here. I only want to
note, in connection with the interpretation of Kant, two uncontroversial facts. First,
among the examples of motions that Aristotle gives — losing weight, walking,
building, and so forth — he also includes learning (manthanein). For learning, as the
acquisition of a new cognitive capacity, is an alteration in the soul. (This does not
mean that learning is not in the soul’s nature. Certain motions can, presumably, bewhole corpus. Its particular connection with the Metaphysics is not important for the points I want tomake. See Burnyeat (2008), 219-92.37Kant makes a point of distinguishing alteration (Veränderung) from change (Wechsel) at A187/B230f. What is altered (paradigmatically, a substance) survives; what changes (paradigmatically, a state ofa substance) does not.
1. The Formal Modalities 20
natural for a thing.)
And second, Aristotle includes, among his examples of pure acts, knowing.38 It
seems to follow that to know — and, I would add, to judge — is not an alteration in
the understanding, but simply the understanding’s being (more fully) what it is.39
(One indication of this is that — unlike learning — knowing does not presuppose
an opposing state. Someone who knows may never have been ignorant. Someone
who learns must have been.)
2. These considerations have been programmatic, but I think they suggest an
alternative interpretation of the progress represented by the modalities: viz. as pure
acts.40
On this interpretation, the progress from problematic to assertoric to apodeictic
will not, in general, consist in alterations or, in particular, in learning. As I said,
judging is closely connected with learning. It depends upon learning; it is, as
Aristotle might put it, the final cause of learning. But it itself is not learning. This
has two important implications.
First, the progress represented by the modalities will not be intrinsically
temporal. I do not mean that it will occur outside of time altogether. Clearly it does
occur in time in the sense that it occurs under broadly temporal conditions: it
comes to be in time, and is connected with the representation of temporal
phenomena. But the progress itself does not take time.
This can be brought out by considering why motions do take time. Motions
consist in alterations. But alterations, as Kant understands them, are transitions38The terms he uses in Θ.6 (1048b24) are noein and phronein; that suggests that he means the pointto apply to understanding in the broadest sense, both theoretical and practical.39This does not mean that any particular knowledge could not be perfected in innumerable ways.40Kant does not seem to have read much Aristotle, and it is possible that he did not know Meta-physics Θ. I am not claiming, however, that he consciously adopted Aristotle’s approach to knowledge.I am only claiming that the comparison with Aristotle’s approach is instructive. Still, it would besurprising if Kant had hit on Aristotelianism completely by accident. And it is worth rememberingthat in the 18th century, German philosophers frequently drew on concepts from the Greek tradition.Kant would have gotten some knowledge of Aristotle through Leibniz and Wolff and Baumgarten.For a discussion, see Reich (1935), 3 f.
1. The Formal Modalities 21
between opposing states of a thing.41 One state begins to be; the opposing state
ceases to be.42 Because the states are opposed to one another, they cannot coexist.
Consequently, if alterations take place under broadly temporal conditions — if
coexistence means temporal coexistence — they must extend over time. Kant puts it
like this:
[E]very alteration has a cause that shows its causality in the whole timein which it proceeds. Therefore this cause does not bring forth itsalteration abruptly (at once or in one moment) but in time. . .[A208/B253 f., cf. A37/B53 f.]43
With pure acts it is different. Since here the act just is the capacity that acts, it does
not have to take time — even under broadly temporal conditions. (Imagine
someone asking “How long does it take to know that?” All this could mean is:
“How long does it take to learn it?”.)
Now, there can be a considerable lag between the acquisition of a pure capacity
(e.g. a problematic judgment) and the pure act (the assertoric or apodeictic
judgment). The important point, though, is that they can occur simultaneously.
Evidently that is the case with all analytic judgments, which are immediately
grasped as apodeictic, and with what Kant calls inferences of understanding (e.g. the
inference from “Every man is mortal” to “This man is mortal”44). And it seems to
be the case with certain perceptual judgments, e.g. “This cinnabar is red”.45
41Properly speaking, the faculty of knowledge is not a thing in the sense that empirical objects are(cf. A348 ff.). But it is like a thing in a number of respects, as Boyle has pointed out, and for thesepurposes can be treated as such without paralogism. See Boyle ().42This does not rule out that the opposing state should begin to exist in a weak sense as the initialstate comes closer to completion. For example, Aristotle says, “Those who learn [something] at firstdo not yet know [it]. . . for it must become natural for them, and this requires time” (NicomacheanEthics, 1147a22 f.). But alteration can be understood as continuous so long as the extent to which athing is in one state is inversely proportional to the extent to which it is in the opposing state. (Infact that is how Kant thinks of it [see A208/B254].)43Hume held a similar view. See A Treatise of Human Nature ., sec. 2.44Cf. Log secs. 44-55, 09: 115-119.45The same point holds for the moral law: “[I]f I think of a categorical imperative,” Kant says, “Iknow immediately what it contains” (G , 04: 420). Later he refers to this imperative as apodeictic(op. cit., 415).
1. The Formal Modalities 22
Second, the stepwise progress of understanding will not, strictly speaking,
consist in the serial determination of the preceding steps.46 Kant will sometimes use
the word determination loosely to mean any representation of an object, but when he
is being careful, he means something more specific. “A determination,” he says, “is
a predicate that is added onto the concept of the subject and enlarges it”
(A598/B626). For example, in the judgment “Every body is heavy” the concept of
a body is determined by the predicate heavy. By contrast, the pure act of judgment
— an assertion — is not a determination of a problematic judgment. There is
indeed a temptation to speak as though it were: for it seems that something, some
further predicate or predication, must have been added in order to explain the act.
But nothing is — that is precisely the point. (“The actual contains no more than the
merely possible”, is how Kant puts it [A599/B627].47) The only difference between
a problematic judgment and an assertoric judgment is that, while the problematic
judgment is a merely possible assertion, the assertoric judgment is an actual one. It
is a difference that lies entirely in the subject’s understanding of his own act.
3. It is in the nature of a pure capacity to be purely active; for in being purely
active, it is more fully itself, and it is in its nature to be itself. (“There is,” as Leibniz
puts it, “a certain urge for existence or [so to speak] a straining toward existence in
possible things or in possibility or essence itself; in a word, essence in and of itself
strives for existence”.48) In this sense, such a capacity can be called spontaneous, and
Kant will refer to the understanding as a “spontaneity of knowledge” (A51/B75).
This does not mean that every pure capacity always is active. Obviously, not
everyone who can see is always seeing; not all problematic judgments are assertoric
judgments. But — and here is the point — if a capacity is not active this is never
because of something internal to it. Capacities, as such, act. Accordingly, it is never46Irad Kimhi helped me to see this, and many other things.47The remark comes in the specific context of a discussion of rational theology rather than formalmodality, but its significance seems to be completely general.48“De Rerum Originatione”.
1. The Formal Modalities 23
necessary to provide a special explanation for this act. It is enough to say that the
capacity is — a capacity.
Conversely, a special explanation will be necessary for a capacity’s inactivity.
With problematic judgments, this explanation will typically involve a limitation
within the particular subject, something that gets in the way of the natural progress
of knowledge. For example, if someone’s powers of discernment are limited in
relation to his purposes in judging,49 he may acquire the capacity to judge without
a sufficiently clear awareness of having done so. Or — having acquired the capacity
— he may misuse it (see Log sec. , 09: 53). (One type of explanation that
will not be possible here is that the empirical data might be insufficient to warrant
assertion. The acquisition of the capacity presupposes all the conditions necessary
for its act. Otherwise, it would not be a capacity at all, but only one condition of this
act among others.)
With assertoric judgments — now viewed as capacities for apodeictic
judgments — the same sorts of limitations are possible. But there can also be
another kind of limitation, deeper and more difficult to remove — for it belongs to
the nature of human subjects as such. Kant discusses this at length in the
Transcendental Dialectic: the achievement of absolutely apodeictic judgment is not
possible for human beings even in principle, he says, because our knowledge
depends on the vagaries of experience.50 Nonetheless, he thinks that relatively
apodeictic judgments are possible, given some presumptions about the order of
nature.51 (These will include principles of experience, such as the law of cause and
effect, and the laws of nature, such as universal gravitation.)49Kant calls this limitation stupidity (Dummheit — A133 n./B172 n.), but he can hardly have meantit in the ordinary sense. Any finite power of judgment will eventually run up against its limits. (Inthe Logic, Kant himself says that “the charge of absurdity” — meaning stupidity in the ordinary sense— “is always a personal reproof, which must be avoided, especially in the refutation of errors” [, 09: 56].)50I will return to this theme in the next chapter.51One such presumption is the law of infinite specificity in nature — “entium varietates non temereesse minuendas” (A656/B684). There are several others.
1. The Formal Modalities 24
These considerations point to a way of understanding the progress he describes
at A76/B101: namely, as the gradual removal of subjective limitations on the act of
judging. If the limitations are expressed by the word merely, the progress could be
expressed by leaving it off: as a progress from a merely possible judgment to
possible judgment without qualification, and from a merely actual judgment to
actual judgment without qualification.52 That is how (relatively) apodeictic
judgments can be knowledge in a fuller sense than problematic judgments, yet
contain nothing that was not present in them all along.
Conclusion
I have argued that Kant introduces the formal modalities into his theory of logic
not merely to satisfy the requirements of syllogistic, but to prepare the way for a
theory of knowledge in which the understanding makes a certain kind of progress:
a progress from capacity to act, and from that act — viewed as a capacity — to its
act, an actual judgment without qualification. It is a further question why Kant
holds such a theory, which is itself distinctive of the Critical philosophy. In the next
chapter, I will try to answer it.
52As I said, this judgment is still only qualified knowledge; necessary judgment without qualificationis not possible for human subjects.
CHAPTER TWO: THE REAL MODALITIES
Will it not be a fair plea in his defense that it is in the nature of a truelover of knowledge to strive after what exists. . . ?
— Plato, Republic 490a
2.1
1. In Chapter One, I argued that Kant’s conception of formal modality takes its
distinctive shape from the role of the logical modalities in human knowledge. For
this knowledge is essentially a progress from capacity to act, and — I argued — the
formal modalities are logical representations of this progress. But the knowledge
itself is not merely formal. Its content consists in the representation of an object
that exists outside the mind — paradigmatically, a body of some kind.
In section 10 of the Critique of Pure Reason, Kant says that the modality of a
judgment’s content is the concept of this object’s existence as possible or impossible,
actual or non-actual, necessary or contingent — and he calls these concepts the
categories of modality. The purpose of this chapter is to explain their meaning.
2. Kant will speak of existence in two senses.1 In a special sense, it is the concept of
actuality — a category of modality.2 What exists in this sense is not something that
is merely possible, but an object whose possibility has been in some sense realized.
But he will also speak of existence in a broader sense. When he says, for
example, that theoretical knowledge is “that by which I know what exists”1This has largely been overlooked. See, for example, Rosenkoetter (2009); Grier (2001), 258.2In section 10, Kant refers to existence as a modal category. In subsequent sections, however, he
refers only to actuality (A145/B184; A218/B266 et passim). Kahn observes that existentia traditionallycarried this connotation (1972), 142.
2. The Real Modalities 26
(A633/B661), he does not just mean that it is knowledge of what is actual. He
means that it is knowledge of what is in general — i.e. of being, in one of its
acceptations.3
My purpose here is to explain how the two senses of existence are related. I will
begin by saying something about the broad sense, since this is prior not just to the
special sense, but — as I will argue — to all the special categories Kant
distinguishes in the Critique of Pure Reason.
2.2
1. The concept of existence in general is, as I said, closely connected to the concept
of being, but there is a sense of being that does not mean existence. For example:
according to Kant, the judgment “Every body is extended” does not contain the
concept of existence. In some contexts, it may have existential import, since if a
body does exist this will entail that something extended also exists. But in itself, the
is only expresses a relation in the mind between the concept of the subject (body)
and the concept of the predicate (extension), such that the thought of the one
necessarily involves the thought of the other (cf. A595/B623; A598/B626).4
Accordingly, Kant will call this the logical sense of being (A598/B626).
The other sense of being, the one that is of interest here, I will call the existential
sense.5 It is most clearly visible in judgments that contain no additional predicate
— e.g. “A body is” or “There is a body”. But Kant takes the traditional view6 that
it is also found in certain predicative judgments. For example, he thinks that the3Kemp Smith makes this point implicitly by translating “solche. . . , wodurch ich erkenne was da
ist” as “knowledge of what is” (, 526).4Cf. Locke, An Essay Concerning Human Understanding ., sec. 7; Leibniz, New Essays on Human
Understanding ., sec. 14.5Kant does not use this, or any other name for the non-logical sense of being he distinguishes at
A598/B626.6Kahn (1973), 245 ff. finds instances in Homer and Aristotle; Bäck (1987), does not find it in
Aristotle, but does find it in Ibn Sina. Wood (1978), 116 associates it with Caterus.
2. The Real Modalities 27
judgment “Every body is divisible” entails that “A (divisible) body exists”. (I will
explain why below.) For this reason, it is not always obvious which sense of being is
used in a particular judgment; in general, though, it is a question of the judgment’s
content, rather than its logical character.7 I will return to this.
2. Kant gives the concept of being surprisingly little treatment in the Critique of
Pure Reason.8 Nonetheless, a number of passages convey his particular view of it.
The clearest of these, and the best known, is probably A598/B626:
Being [Sein] is obviously not a real predicate, that is, a concept ofsomething that could add to the concept of a thing. It is merely thepositing [Position9] of a thing in itself, or of certain determinations inthemselves.10
Although he does not say which sense of being he means here, it is apparent that
the existential sense is at least in view — for it goes without saying that the logical
sense of being does not add to the concept of a thing, and no one ever thought that
it did.
With this in mind, there are two propositions to consider. The first is:
(R) Existence is not a real predicate.
Traditionally, a real predicate was the concept of an affirmative quality in an
object — a reality, as opposed to a negation. This, for example, is from
Baumgarten’s Metaphysica, the textbook that Kant used for his regular metaphysics
course:7When I speak of logic, I will always mean what Kant calls general logic, “the science of the rules
of the understanding in general” (A52/B76).8Perhaps this is because — as he says in — “this concept is so simple that nothing can be said
to explicate it. . . ” ( 02, 73).9In other contexts, Kant will use Setzen or Setzung rather than Position. This does not seem to
correspond to a conceptual distinction.10Similar passages can be found in other works — for example, in ( 02, 73), and in the firstintroduction to the Critique of Judgment ( 20, 402). See also: R 3724, 17, 263 (ca. 1763); R 4298, 27, 499 f. (1770s); R 4030, 17, 390 (1769 or 1770); R 4396, 17, 531 (ca. 1770s); R 3761, 17, 286 (1760s); , 22, 549.
2. The Real Modalities 28
What is posited in something to be determined (marks and predicates)are determinations. Either a determination is positive and affirmative —which, if it is truly [affirmative], is a reality; or negative — which, if it istruly [negative], is a negation. [Sec. 36. See secs. 135 ff., sec. 248. Cf. 17, 34.]
Such predicates were thought to be of great methodological importance, and were
applied across a wide range of domains — including not only physics but also
moral and speculative philosophy (thus God was called ens realissimum, the most
real being [A576/B604]. Cf. Baumgarten, Metaphysica, sec. 807).
Precisely because the range of these predicates was so broad, however, there
could be no standard procedure for deciding whether a given predicate P was real.
In particular, this could not be decided by appealing to affirmative judgments with
P as predicate (e.g. “Every body is P”). For — as Baumgarten himself saw — the
question was not whether P had a certain kind of logical character, but whether it
represented something in reality. This was a metaphysical question. The answer
was necessarily a metaphysical claim.11
Now I take it that R is a claim of this kind, though a negative one: it is the
proposition that the concept of existence does not represent something in reality.
Unfortunately, there is no accompanying proof; and though Kant makes several
interesting remarks following A598/B626 (including the famous one, that “a
hundred actual thalers contain not the least bit more than a hundred possible”),
these are not conclusive. For this reason, Allen Wood has suggested that he simply
“regards its truth as something obvious, something ‘every rational person admits’”
(1978), 107.11See .. Meier, Metaphysik, sec. 48:[T]o find out whether a determination consists in a true addition, or [only] a privation[Mangel], one must neither be misled by words, nor by a superficial [ersten] aspect, butmust penetrate more deeply into its nature.
Anneliese Maier, who cites this passage, points out that there is no evidence that Kant had read it(1930), 65. She also points out, however, that the remark is typical of the metaphysics of the earlymodern period, and the point was surely familiar to him Cf. Locke, Essay ., secs. 1-6.
2. The Real Modalities 29
Clearly, there is something right about this. Kant himself says that R is
“obvious”.12 But this does not mean he thinks it is obvious apart from any other
ontological commitments. And in fact there are a lot of these in the Critique of Pure
Reason. (Some of them are the subject of the present work.) I want to suggest that
R can only be understood in relation to them.
3. The second proposition is:
(T) Existence is the positing of a thing in itself,or of certain determinations in themselves.
Positing is not defined in the Critique of Pure Reason. Kant may have thought its
meaning would be obvious to his readers.13 Or maybe he thought he already had
defined it: in an essay called “The Only Possible Ground of Proof for a
Demonstration of God’s Existence”, he says that “the concept of positing [Position
oder Setzung] is . . . the same as that of being in general”.14 In any case, he seems to
have something like this in mind at A598/B626.
If that is right, then positing something means representing that it somehow is.
And since there are two senses in which something is said to be, there will also be
two senses in which something is said to be posited. In the logical sense, what is
posited is a relation between mere concepts (e.g. body and extension); in the
existential sense, what is posited is the thing in itself.15
Now, in speaking of a thing in itself, Kant often means something special: the12He does not actually say it is something that “every rational person admits" — that quotation isfrom a different context. But Wood is right that this is what he thinks.13Positio and its cognates posito and ponere were common terms of Latin philosophy. (They canalready be found in Burley’s 1302 treatise, De Obligationibus, secs. 3.01 ff.) Kant would have knownthem from Wolff (e.g. Logica, sec. 406) and Baumgarten (e.g. Metaphysica, secs. 12 ff., et passim)among others.14, 02, 73.15Wood says thatKant’s view about the ‘positing’ function of the copulative [sc. ‘logical’] ‘is’ seems tocommit him to [the] view that every true predication presupposes the actual existence ofthat to which the subject term refers. [1978, 116.]
This is not right, however, for Kant clearly recognizes a merely logical sense of positing. See Zöller(2008).
2. The Real Modalities 30
supersensible correlate of sensible objects, about which (he notoriously thinks)
nothing, or almost nothing, can be known (B f. et passim). That is not what he
means at A598/B626, however, and it is not how I will use the term.16 The thing
that is posited is paradigmatically an object of experience.17 What makes it a thing
in itself is simply that — as an object of experience — it exists independent of its
concept in the mind.18
I should add two qualifications.
First, the thing in itself is not absolutely independent of its concept in the mind.
Kant holds that in general, the objects of human understanding “must conform to
our knowledge”, in the sense that the “representation is a priori determining of”
their form or essence (B; A92/B125).19 When I say that the thing in itself is
independent of its concept, then, I do not mean that everything that can be known
about it can be known from experience. I only mean that the fact that it exists does
not follow from its concept. For this is an empirical fact. As he puts it:
In the mere concept of a thing no mark [Charakter] of its existence is metwith at all. For although it may be so complete that not the least bit islacking. . . existence has nothing to do with all this, but only with thequestion: whether a thing is given to us in such a way that theperception could precede the concept [i.e. empirically]. [A225/B272. Cf.R 5710, 18, 332 (ca. 1780s); R 5772, 18, 349 f. (ca. 1780s); R 6413, 18, 708 (1790s).]
Second, because the thing in itself exists independent of its concept, the
representation of this existence will only be possible under certain conditions, and16In the Transcendental Aesthetic, Kant points out that the phrase thing in itself can be used intwo senses: an absolute sense, in which it denotes the supersensible; and a relative sense in whichit denotes an empirical object (what he calls an appearance, Erscheinung) in contrast to an empiricalappearance (what he calls an illusion, Schein). For example, he says, in relation to the rainbow thatappears in a sunny shower, the rain is the thing in itself (A45/B63).17This is confirmed by a Reflexion from 1790 — after the second edition of the Critique had been
published — in which Kant speaks explicitly of “something outside us” in space (thus a sensibleobject) as a “thing in itself” (Sache an sich) (R 6312, 18, 612).18See A639/B667: “The knowledge of the existence of the object consists in its being posited in itself,outside thought”; and R 6232, 18, 643 (1793), where he speaks of “what exists not merely in myrepresentation (but (as thing) in itself)”. It was a common usage. Cf. Locke, Essay ., sec. 1; .,secs. 8, 23; ., sec. 2 et passim.19This is the celebrated Copernican Hypothesis. Cf. A79/B104 f.; A158/B197.
2. The Real Modalities 31
subject to certain rules. Kant’s word for such a representation is determinate
(bestimmt): to know the existence of an object, he will say, is to have a determinate
representation of it. I will try to explain what this means.
2.3
1. According to Kant, a representation is determinate when a certain kind of
predicate has been added to it (A598/B626; Log sec. 36, 09: 111). He
emphasizes that this is an addition to the content of representation: the concept of a
divisible body has more in it than the concept of a body alone; and the concept of a
heavy divisible body more in it still. But it also pertains to the form of
representation: the combination of these concepts is not just another concept. It is
also a certain kind of judgment.20
2. This judgment is the act — or as Kant will say, the function — of the
understanding (A70/B95 et passim). In the first place, he will call it a logical act, and
the function a logical function: for in the first place a judgment is simply a mental
representation that is ordered according to certain logical rules.
In section nine of the Critique, the logical function of the understanding is
divided into a number of sub-functions, or moments, corresponding to the forms of
judgment identified by traditional logic. (The so-called Table of Judgments
presented in that section is, more properly, a table of logical sub-functions and
sub-sub-functions.)
Of these, two are primary.21
20I use the word also advisedly. For Kant a logical judgment is simply a special kind of concept.See Log sec. 17, 09: 101.21In fact, Kant lists the logical functions of judgment under four headings (A70/B95). But the twofunctions of quantity and quality— though they are very important — are not primary functions. Theirrole is subordinate to the relation of concepts contained in a categorical judgment; they do not evenapply in the case of hypothetical or disjunctive judgments, except indirectly.
2. The Real Modalities 32
a. In a judgment, the understanding effects a certain relation among concepts
(A70/B95). In the basic case, it is a relation between a subject-concept (e.g. body)
and a predicate-concept (e.g. divisibility). Kant will characterize this relation in
different ways in different works,22 but the underlying idea is always that the
predicate is asserted of the subject; it says something about it (A322/B378; cf.
A69/B94). Accordingly, this relation is called categorical — as in “Every body is
divisible”.23
Although a categorical judgment is a logically complete representation, it can
also be related to another such judgment through a higher-order function. The
latter is either hypothetical (i.e. conditional) or disjunctive (in the exclusive sense)
(Log secs. 23 ff., 09: 104 ff.; A73 f./B98 f.). (These relations are extremely
important to Kant, but for the sake of simplicity I will set them aside here.)
b. The understanding also effects a certain consciousness of itself in judging (Log sec.
17, 09: 101; B141). This consciousness does not add anything to the content of
the judgment; it does not introduce a further concept. Nonetheless, it is essential to
the function of the understanding in judgment; for it is the consciousness that I am
judging that makes a mental representation a judgment at all, as opposed to a mere
thought (which may share its content [B142]). Accordingly, this is the fundamental
logical function — and the condition of all the others.
The logical value of self-consciousness in a judgment is the judgment’s formal
modality (A74/B99 f.; Log sec. 30, 09: 108 f.). And, as I explained in Chapter
One, Kant divides this into three special functions — the problematic (if the
judgment is regarded as merely possible), the assertoric (if it is regarded as actual),
and the apodeictic (if it is regarded as a necessary assertion).22In particular, there seem to be discrepancies between the descriptions of predication in the Critiqueof Pure Reason and the Logic. I will not try to reconcile these here; for an interesting attempt, seeLonguenesse (1998), 86 ff.23The Greek word kategoria means something said.
2. The Real Modalities 33
3. The two primary logical functions of the understanding are conditions of the
determinacy of representations. Determination itself, however, is not a logical
function: for the determining predicate adds something to the content of a
representation, while — according to Kant — logic abstracts from all content
(A54/B78 et passim). Determination is, rather, the real function of judging (as I will
call it).24 I will try to explain what this means.
a. The content of a judgment consists in the concepts it comprises. For example, in
the judgment “Every body is divisible”, the content consists in the concepts of body
and divisibility. This content is logically necessary, since a judgment must be about
something or other; even at the highest level of abstraction, one cannot judge that
“Every is ”. But the content is not there to satisfy a logical requirement. In
fact, logic does not require that there be any judgments at all; it only provides the
formal conditions under which judgments are possible. The content is there, at least
in paradigm cases, to bring the judgment into relation with an object: a body, some
metal, things in general, or whatever.25
It is the representation of this object that is determined in the act of knowledge
as Kant understands it (see B137). And it is this representation that, in the act of
theoretical knowledge, is the representation of a thing in itself.26
b. A judgment represents an object through its content. A judgment represents an
object determinately just in case this content is determinate.24I do not mean real in the sense of genuine, but in a sense related to that of 2.2 above — the sensein which one can speak of a real predicate. Although Kant does not speak of real functions as such, hefrequently contrasts the logical use of a function with its real use (e.g. at A299/B355).25Not every judgment is about an object. A judgment of perception is merely “a connection ofperceptions in my mental state”; the judgment that “Sugar is sweet”, for example, “only express[es] arelation of two sensations in the same subject — namely, myself — and even then only in my presentstate of perception. . . ” (Prol. sec. 20, 04: 300; sec. 19, 04: 299). Clearly, though, it is not theprimary purpose of content to express “my present state of perception”. The content I will speak ofhere is assumed to bear some relation to an object.26With practical knowledge things are considerably more complicated, since the object is not reallydistinct from the representation of it. Unfortunately, there is not space to address these complicationshere.
2. The Real Modalities 34
But content can be called determinate in two senses. In a loose sense, it is
enough that it is determined in some judgment or other. For example, the judgment
“Every body is extended” has determinate content in the sense that there are
determinations that involve the concepts of body and extension.
In a stricter sense, though, the judgment does not have determinate content, for
it does not contain a determination. According to Kant, the concept of a body
already contains the concept of extension; body that is not extended is a contradictio in
adjecto (cf. Prol. sec. 2, 04: 267). The judgment simply makes this explicit — it
does not introduce new content into the understanding. For this reason, Kant will
call it merely explicative or analytic [A6 f./B10 f.; cf. A150 ff./B189 ff.].)27
A judgment that does contain a determination is called synthetic, because it
represents a combination (or synthesis) of logically distinct representations (A6
f./B10 f.). (Kant will also call this judgment ampliative [Erweiterungsurteil], because
the combination adds new content.) In doing so, it makes a claim about how things
really are. For example, in the judgment “Every body is divisible”, the claim is not
simply that in thinking of a body, one thereby thinks of something divisible. It is
that divisibility really belongs to bodies — to the things themselves (see B142).
To make such claims is the real function of a judgment. Now I want to say a
little about what this involves.
4. In the first place, the content of a synthetic judgment consists in a number of
logically distinct concepts. For example, the content of the judgment “Every body is
divisible” consists in the concepts of body and divisibility, neither of which logically
contains the other.28 But — Kant says — so far
it remains undetermined which of the two concepts will be given the27Analytic judgments can have a role in the determination of content. For example, the judgmentthat bodies are extended may make it easier to see that they have a certain determination. But thisrole is not represented in the judgment itself, even implicitly. See A10/B13 f.28Kant says that this judgment is synthetic at B128; cf. A68/B93. It is not important whether he isright, so long as some judgments are synthetic.
2. The Real Modalities 35
function of the subject, and which that of the predicate. For one can alsosay: “Some divisible is a body”.29 [B128 f.]
It is an important remark, though a difficult one. I take it he is not denying
that, in the judgment “Every body is divisible”, the concept of a body has the
function of a subject, while the concept of divisibility has the function of a predicate.
His point is rather that these concepts have not yet been assigned definite functions
in judgments in general. “For one can also say: ‘Some divisible is a body’.”
It can be hard to see why this matters. There is certainly no logical
requirement that a concept have the same function in every judgment. Indeed, if
the inference from “Every body is divisible” to “Some divisible is a body” is valid
(as he believes30) then the concepts of a body and of divisibility must be able to
convert their logical functions.
But for Kant there are two aspects of a judgment, a form and a content. And
the point of the remark seems to be that — while the conversion is formally valid —
the content of the judgment “Every body is divisible” is not convertible.
a. A determination adds something to the content of the subject-concept. (In the
example, the concept of divisibility is added to the concept of a body.) Kant finds it
natural to speak of addition here because the judgment makes a real contribution to
knowledge — it does not just clarify things. But this can be misleading. Addition, at
least in the common understanding of it, is a commutative relation: if M+N = Σ,
then N+M = Σ. But the relation that the judgment represents between the concept29The word I have translated as divisible is a substantive adjective (Teilbare). In English this wouldnormally be divisible thing, as in “Something divisible is a body” (cf. , 128). I have retained theliteral sense, however, to emphasize that the two judgments have the same content.30Logical conversions are discussed at Log secs. 51 ff.:Immediate inferences through conversion [Umkehrung] concern a relation of judgments,and consist in the transposition of the subjects and the predicates in the two judgments,so that the subject of the one judgment is made the predicate of the other judgment, andconversely. [ 09: 118.]
With universal affirmative judgments, logical conversion also affects the quantity (and so is calledaltered conversion, or conversio per accidens — cf. Prior Analytics .2, 25a1 ff.). That is why theconverted judgment is particular, even if it also happens to be true that every divisible is a body.
2. The Real Modalities 36
of divisibility and the concept of body is not commutative in this way. It makes a
great difference whether one says “a body is divisible” or “a divisible is a body” —
even if both are true in some sense.31 This was already implicit in the logical
function of relation, which assigned differential roles to the subject-concept and the
predicate. The point here, however, is that this same function can be exercised in
relation to the content of judgment, and so in relation to real objects.32 This is what
makes it possible to represent a relation between something that is in itself a subject
(e.g. a body) and something that is in itself a determination (e.g. divisibility).
Now, Kant holds that the concept of something that is in itself a subject is the
category of substance; the concept of something that is in itself a determination is the
category of accident (A246). A lot could be said about the categories, and their role
in the Critique of Pure Reason; here, however, I will only mention a few points.
b. Just as he will speak of the logical function of judgment, Kant will speak of the
category, meaning the concept of a determination as such (B144 et passim). The
interrelated moments of the category are the special categories.
A complete table of these categories is given in section 10 of the Critique. Since
the category is an exercise of the very same function that effected the logical form of
judgment, each of the special categories corresponds to a logical sub-function (as,31How can a logical conversion be true if content is not convertible? I cannot give a comprehensiveanswer to this question here, but the basic idea is that the truth of the conversion will depend on thecontent of the original judgment. In the example I have been discussing, it is true that some divisibleis a body because it is true that every body is divisible — and not the other way around. See Prol.sec. 2:
[A] synthetic proposition can certainly be discerned [eingesehen] according to the principle[Satz] of contradiction, but only if another synthetic proposition is presupposed fromwhich it can be inferred, and never in itself. [ 04: 268.]
32See A79/B105:The same function which gives unity to different representations in a judgment also givesunity to the mere synthesis of different representations in an intuition [i.e. a given ob-ject]. . . The same understanding, therefore — and indeed through the very same actswhereby it brought forth the logical form of a judgment. . . also introduces a transcen-dental content into its representations. . . which concerns objects a priori. . .
2. The Real Modalities 37
for example, the categories of substance and accident correspond to the categorical
function of judgment).33 For the same reason, the primary division is between the
categories of relation and the categories of modality. The former pertain to the
content of judgments directly; the latter pertain to the consciousness constitutive of
this content.34
c. The addition of categories invests the particular content of a judgment with a
kind of metaphysical significance. It is not just that familiar things can now be put
into the terms of classical metaphysics (so that, e.g., “Every body is divisible” will
now express the inherence of an accident in a substance).35 It is that judgments
about such things can now be seen to represent what is, in the deep sense that
classical metaphysicians had sought to describe: substances, accidents, causes,
effects, possibilities, actualities, necessities, contingencies.36 And this is exactly what
Kant says: the categories in general are concepts of “the existence of. . . objects”
(B110, emphasis added). The relational categories are concepts of the existence of
objects “in relation to one another”; the modal categories are concepts of the
existence of objects “in relation to the understanding” (ibid.).
5. As the concept of the determination of an object, the category is the same as the
concept of existence in the general sense. This is the outcome of the previous
subsection. Now I want to add that — because this determination is the real
function of judgment — what exists in the general sense is simply the object of a
determinate judgment.37 Which is to say — as Kant does — that knowledge is33See the Appendix for a table of correspondences.34See A219/B266: “The categories of modality have this peculiarity: that in the determination ofthe object they do not in the least enlarge the concept to which they are added as predicate, but onlyexpress its relation to the faculty of knowledge.”35The advantage of putting things this way can be hard to see today, but for Kant it would havebeen obvious: one of his main purposes was to give a proper treatment of philosophia perennis. See,for example, G, 04: 387 f.36Metaphysics, Aristotle says, is the science “that contemplates being as being” (Metaphysics Γ,1002a21). For him — and for Kant too — the categories are the primary terms of this science (cf.A79 f./B105).37Mathematical judgments may be an exception, since they are determinate, yet their objects do not
2. The Real Modalities 38
about what there is.38
6. To explain how these points bear on existence in the special sense, i.e. that of
actuality, it will first be helpful to consider an alternative interpretation of the way
existence in the general sense fits into Kant’s theory of judgment — one that has
attracted a number of recent commentators.
2.4
1. Kant does not think that every synthetic judgment contains a determination. For
example, he thinks that the judgment “There is a body” is synthetic (A598/B626),
but he does not think that the concept of a body is determined in it.
This has led some commentators to conclude that there must be two real
functions of judging: first, to represent a determinate relation, and second to
represent the existence of the object these contents represent. Here is how Allison
puts it:
[W]hat makes any judgment synthetic is that it “materially” extends ourknowledge beyond what is already thought (implicitly or explicitly) inthe concept of its subject. But it can do this in either of two distinctways: by affirming (or denying) a further determination of the subject orby affirming (or denying) that the concept of the subject is instantiated.Existential judgments are synthetic in the latter way.39
And here is Wayne Martin:seem to exist in the ordinary sense. It is worth noting, however, that — for Kant — these judgmentshave determinate content only on the assumption that “there are things that can be presented to usin accordance with” mathematical principles, and to which mathematics can therefore be applied(B147, emphasis added). A major purpose of the Critique of Pure Reason is, of course, to establish thisapplicability (see A165/B206). (For illuminating discussions, see Sutherland (2004) and (2005).)38I have put things in terms of Kant’s definition of theoretical knowledge, since this is the mainconcern of the Critique, but it is worth noting that the point is not limited to this knowledge. Prac-tical knowledge, according to Kant, is “that whereby I represent to myself what should exist” (seeA633/B661). Since what should exist is at least possible — since ought implies can— practical knowl-edge is also knowledge of existence in the general sense. (See A548/B576, A807/B835; , 06:380; Rel, 06: 47, 50, 62, 64; KpV, 05: 142 f.)39(2004), 414 f.
2. The Real Modalities 39
For the purposes of inference, existence can be treated as a concept (eithersubject or predicate) and existential judgment can be modeled in thelogic of synthesis [i.e. as ordinary synthetic judgment]. But while thistreatment suffices for syllogistic proof, it fails to capture the form ofexistential judgment as a distinctive mode of thought in its own right — as aform of positing or thesis rather than of synthesis or combination. 40
Allison and Martin differ in many particulars, but their interpretation shares an
underlying idea: namely, that Kant at least implicitly recognizes a division of
cognitive labor between determining a concept and representing something existing
under it.
Although this is not completely wrong — Kant certainly does distinguish
between real functions of relation and modality — it cannot be completely right
either: for the relation of concepts in a determination is already a representation of
existence. That is the meaning of the category, after all. As Kant himself points out,
“The determinations of a substance are nothing other than the particular ways in
which it exists” (A186/B229).
2. The right thing to say, I think, is this.
There are two factors in a determination. There is (1) the content, which is
specified by the categories of relation; and there is (2) the consciousness of this
content, which is specified by the categories of modality. Although both factors are
essential for knowledge, they are essential in different ways. The relational
categories are essential for the determinacy of judgments, since they alone make it
possible to represent real predications. The modal categories, however, are essential
for the judgment to be a judgment at all, since they represent the consciousness that
constitutes it as such.
It follows that, while the relation contained in a determination depends on the
modality — since without this it would not be knowledge at all — the modality
does not depend on the relation in the same way. A judgment that did not contain40(2006), 55 (emphasis added). Cf. Anderson (2008), 105 n.
2. The Real Modalities 40
any relation at all would be without determinate content. But it would still be an
exercise of a real function of judgment, and so, would still be knowledge in an
attenuated sense.41
I take it that this is the case with a judgment like “There is a body”. Here body
is not a subject that is related to a (real) predicate; it does not even come under the
category of substance.42 The judgment is simply a modal representation of a
particular concept: the representation, in a judgment, of the kind of consciousness
characteristic of representations in knowledge. In other words, it is the
representation that the concept figures in knowledge — without, however,
determining the object of this knowledge. (Which is just what it means that existence
is not a real predicate.)
2.5
1. Now that I have said something about what it means to represent a concept
under a modality in general, I want to say something about existence in the sense of
actuality.
2. In the first place, special existence, or actuality, is a category of modality — a41Kant does not say this explicitly, but he does make an analogous point about concepts, which Iam paraphrasing. Without content, he says, a concept “would [still] be a thought, formally speaking[.]It would, however, be without any object, and through it no knowledge of anything would bepossible. . . ” (B146).42An alternative would be to interpret “There is a body” as a disguised predication. There are twoways of doing this.
(1) “A body is existent.” Then body would be a substance, and existence an accident. This seemsunpromising, however. An accident is a way of existing. But existing is not a way of existing.
(2) “An existent is a body.” Kant himself had proposed this interpretation in an early work(, 02: 74), and unlike (1) it is not obviously wrong. The idea would be that existenceis a substance, and everything else an accident. (As Martin notes (2006), 55, this looks likeSpinozism.) Kant’s specific reasons for abandoning this idea — which he must have donein the 1770s — are not known. Presumably, though, they are connected with his eventualconception of nature as a community of reciprocally limiting finite substances (see the ThirdAnalogy, A211 ff./B256 ff.). For if substances can limit one another, there cannot just be one ofthem (i.e. existence).
2. The Real Modalities 41
concept of the manner (or mode) in which content is represented in knowledge
(hence modality). In the Critique of Pure Reason, Kant identifies six such concepts,43
and presents them in a table at A80/B106 — the Table of Categories:
.
Categories of ModalityPossibility – ImpossibilityExistence – Non-being [Nichtsein][Actuality – Non-actuality]Necessity – Contingency.
Everything about this table – the arrangement of concepts, the pairing of
contradictories, the symmetrical em dashes, the pattern of textual emphasis —
everything points to an underlying principle. The question is what that is.
3. I cannot give a full answer here – partly because Kant does not say very much
about it, and partly because the systematic division of categories in general involves
special complications, which would require special treatment. But – given the
account of the functions of modality I developed in Chapter One – I can make some
preliminary remarks.
a. Earlier I said that – for Kant – it is always an empirical question whether
something exists.44 (Indeed, he says, “If we try to think existence through the pure
category alone. . . we can point to no mark that distinguishes it from mere
possibility” [A601/B629]). Even a priori knowledge, he thinks, must bear a relation
to experience, for it is a special kind of knowledge of existence.4543According to the Prolegomena, there are only three categories of modality — possibility, existenceand necessity (sec. 21, : 04, 303). Clearly these are the primary categories in any case.44See sec. 1.2.3, above.45See B294:
2. The Real Modalities 42
This dependence on experience is characteristic of finite understanding. An
infinite understanding, such as Kant imputes to God, would know everything
immediately; the question whether something existed could not arise for it, even in
principle. (At B138 f., for example, Kant characterizes God’s understanding as one
“through whose representation the object of this representation would at the same
time exist. . . ” [emphasis added].) By contrast, a finite understanding – such as
human beings have – can only know that objects exist when the mind is affected by
them in experience. To be sure, the representation of this affection is not by itself
knowledge of anything, since it is not conceptually articulate. It is, however, an
essential condition of knowledge; for it is only in being affected that the
understanding can “work up” the knowledge of what exists (B1).
The working up is, of course, an essential part of the real function of human
understanding. But this implies that – in such an understanding – this function is,
by its very nature, subject to limitations. For it cannot be exercised immediately, in
isolation from sensible material; it takes work!46
Now, in Chapter One, I said that the limitations on the merely logical function
of modality were associated with its division into the three moments of problematic,
assertoric and apodeictic judgment. I also said that these limitations were associated
with Kant’s broader theory of knowledge; now I can explain why. It is because
human understanding can only know its object by working up sensible material
that it must progress in its knowledge from possibility to actuality to necessity. The
stages in this progress correspond to stages in its work toward self-understanding.
As I also said in Chapter One, it is important not to confuse this kind of
progress with the process of learning. It is true that human understanding mustAll principles [Grundsätze] of the pure understanding are nothing but principles [Prinzip-ien] a priori of the possibility of experience, and all synthetic propositions a priori relateto this alone – indeed their very possibility rests entirely on this relation.
Cf. A156 ff./B195 ff.; A395.46The word function already suggests this: it comes from a Latin translation of the Greek wordergon, which is itself the source of the English word work.
2. The Real Modalities 43
learn what it knows; and this is necessary because of the very limitations I
described. Learning is therefore one kind of progress in understanding. But it is not
the only kind. Even when it has the data it needs, and so is in a position to know an
object, the work of this understanding is not done – for it still may not know that it
can know this object. It follows that there is room, within the act of knowledge
itself, for a distinction between the possibility of knowing and its actuality; and
between the mere actuality of knowing and the knowledge of its necessity.47
This distinction is of course merely subjective: it does not make any difference
in the content of knowledge. (There are no possible objects or actual objects or
necessary objects, for Kant – only possible and actual and necessary knowledge of
those objects.) But this is not a merely psychological distinction, one that arises
from empirical facts about the state of a particular individual’s mind at a particular
point in time. The distinction belongs to the very nature of human knowledge, as
the knowledge of a limited understanding.
Conclusion
1. The relation between the two senses of existence can now be put like this. The
category of actuality is a representation of existence in the general sense, so far as the
latter is (1) related to the understanding and (2) subject to the empirical conditions
under which human knowledge is alone possible. Under these conditions, existential
knowledge in general — once it is sufficiently understood — will be the same as actual
knowledge. But because it is not given that all such knowledge is sufficiently
understood — they are not necessarily the same. There can be merely possible
knowledge. And the knowledge that something is merely possible is nothing more
than the recognition of the work that remains to be done for us to know it.47As Aristotle says, “Those who first learn something. . . do not yet know it; for it has to grow intothem” (Nichomachean Ethics 7.3, 1147a21 f.).
2. The Real Modalities 44
2. In the remaining chapters, I will try to explain Kant’s account of this work in
greater detail. It will require both an understanding of the particular shape of
human experience and of the principles under which modal concepts can be applied
in it.
PART TWO: MODALITY AND THE DOCTRINE OF JUDGMENT
CHAPTER THREE: THE MODAL SCHEMATA
Think’st thou existence doth depend on time?— Byron, Manfred.
3.1
1. The English word judgment can be used in two senses. In one sense it refers to
an act of the understanding; this is the sense I have had in mind in Part One, the
sense in which (for example) “Every body is divisible” is a judgment.
In another sense, though, judgment refers to a power of the understanding:
something like discernment. It is in this sense that someone is said to have good
judgment — that is, to be intelligent or even wise. (By contrast, acts of judgment
are not normally said to be good or bad, but only true or false.)
Although the two senses of judgment are very different — and Kant himself
will use different words for them1 — they are obviously connected. The quality of
our judgments depends in large part on the quality of our judgment; deficiency in
the latter is a principal cause of error in the former. For — as I indicated in
Chapter Two — the act of judgment is not just a logical relation of concepts, but
also a use of them in relation to our experience. The power of judgment, as Kant
conceives of it, is the faculty that governs this use under the sensible conditions in
which human experience is alone possible.2 It is the subject of Part Two.1The first sense he calls Urteil; the second he calls Urteilskraft, i.e. the power to judge. German
requires this distinction: Urteil cannot refer to a power; Urteilskraft cannot refer to an act.2Because the function of the power of judgment (at least in relation to theoretical knowledge)
is the same as that of the understanding — namely, to judge — the former may be regarded as aspecification of the latter: judgment is the understanding, so far as it can act under sensible conditions(A131/B169). For a brief but illuminating discussion of Kant’s conception of the difference betweenfaculty and power, see Longuenesse (1998), 7.
3. The Modal Schemata 47
2. The sensible conditions of experience may be divided into two classes.
a. Those associated with the content of experience I will call empirical. They
comprise broad features of the natural world, and the physiology of human
perception; but they also include quite specific facts — for example that Caius is a
man.
It will be convenient to call the power of judgment, so far as it can act under
such conditions, empirical.
b. The sensible conditions associated with the mere form of experience I will call
pure. These too can be characterized at different levels of specificity; the most
general condition, however, is the representation of time.3 Empirical
representations are thus temporal representations; between any two events in time,
there is always a relation of precedence and succession or simultaneity.4
3. As a pure condition of experience, time is like the categories. But it is not a
category, for it does not represent existence as such; time itself does not exist.5 It is
merely a sensible condition of experience, a form of receptivity to representation
(A22 f./B37).6
Time is not a category, and the categories themselves have no temporal content.
Nonetheless, Kant thinks, they can only be used in relation to time. For (as I said in
Chapter Two) all our judgments — including judgments a priori — must relate to
experience; and experience is temporal. It follows that existence as it relates to our
knowledge is always temporal existence.
The power of judgment, so far as it merely governs this relation, may be called
pure.3Kant presents several arguments for this in the Transcendental Aesthetic (A30 f./B46 f.), but this
is not the place to consider them.4For an extended critical treatment of this point, see Rödl (2005).5I mean existence in the general sense I described in Chapter Two. I will return to this point in
3.3.4, below.6The association is famously qualified at B161 n., but the fundamental point remains the same.
3. The Modal Schemata 48
4. Even a comparatively full account of human knowledge, such as Kant aims to
give in the Critique of Pure Reason, will say little about the empirical power of
judgment. It can give rules of thumb — for example, that one should not judge
hastily or harbor prejudices, and Kant will sometimes mention such things.7 In the
end, though, empirical judgment must be sensitive to conditions that — just because
of their particularity — cannot be anticipated in general propositions; it is more a
matter of practice than of theory.8
It is different, however, with the pure power of judgment. Because the pure
condition of its use (i.e. time) can be represented a priori, it can be treated in an
account of human knowledge. And in fact, it must be treated in the account, for
otherwise we could not explain how experience is possible under the conditions
peculiar to our sensibility — or more specifically, how sensibility matters to the
modality of knowledge.
Kant calls this treatment the Transcendental Doctrine of Judgment9 (A136/B175).
It comprises the central parts of the Analytic of Principles, and arguably of the
whole Transcendental Analytic.
5. I am not yet ready to discuss it. The question of how to use the categories in
relation to time only makes sense if they can be used that way. But this cannot just
be assumed, for the categories themselves have no temporal content; it must be
explained.
The explanation is the topic of the present chapter. As before, I will mainly be
concerned with the modal categories, though some points will have broader
application.
The Doctrine itself, and the principles of modality that are brought to bear in7See for example KpV, 05: 9 n.; Log sec. , 09: 75 ff.; contrast, however, sec. , 09: 18 f.8Cf. A133/B172: “[W]hile the understanding is capable of being taught and equipped with rules,
judgment is a special talent that cannot be taught but only practiced” (cf. /, 8: 275 ff.). It was atraditional view, as the Scholastic maxim — de particularibus philosophia esse non poterit — attests.9It is called transcendental because it pertains to the pure power of judgment (cf. A11 f.; B25); and
it is called a doctrine because it contains instructions for the proper use of this power.
3. The Modal Schemata 49
it, are discussed in Chapter Four.10
3.2
1. The representation that makes possible the use of a concept under sensible
conditions is called the schema of that concept. (The representation of a concept
through its schema is called its schematism.) It is not clear whether Kant thinks the
schema is a kind of concept, or a form of representation in its own right.11 It is
clear, however, that schemata, like concepts, have an aspect of generality: they make
concepts applicable across a range of sensible conditions (A140/B179).12 For
example, if the concept of a tree is the representation of a plant with certain
distinctive marks (e.g. a long, woody trunk), the schema for this concept will be the
capacity to represent trees under empirical conditions.
Empirical schemata normally operate in obscurity. For example, we are not
normally aware of the procedure of our understanding in representing trees. It can
of course be clarified in certain ways, as in a practicum on dendrology. Even then,
however, it will not be perfectly clear — for the empirical conditions of its operation
are not fully specifiable.
2. By contrast, the pure (or transcendental) schemata, so far as they concern the use
of the categories merely in relation to time, can be made perfectly clear. For, as I
argued in Chapter Two, the categories are concepts of existence as such; the10Once again, I am following Kant’s own division of the topic. See A136/B175.11Some commentators have held that a schema could not be a concept. I do not want to argue
with them here; I only want to note that in at least one place Kant refers to the schema as “thephenomenon, or the sensible concept of an object, in agreement with the category” (A146/B186). (Acorresponding ambiguity attaches to the whole faculty of schematism – i.e. the imagination – whichKant calls a “function of the soul” at A78/B103, but which, in his own copy of A, was emended to “afunction of the understanding” [E , 23: 45, emphasis added].)12A number of commentators have identified schemata with “recognitional capacities”, i.e. capacitiesto recognize instances of a concept as such. It is true that to have a schema is to be able to recognizeinstances of a concept. But the schema is also supposed to constitute a concept’s meaning (Bedeutung).And this suggests that it is a condition of the possibility of the concept’s having instances at all (cf.A139/B178; A146 f./B185 ff.).
3. The Modal Schemata 50
schemata for the categories are therefore nothing but capacities to represent existence
under the pure condition of experience, or in other words, capacities to represent
temporal existence.13
The schemata for the special categories are characterized more or less
mechanically, and Kant does not bother with a transcendental deduction of their
validity;14 he merely lists them — or rather most of them — between A142/B182
and A145/B184. Despite its terse presentation, however, the list makes significant
contributions to the account of knowledge.
For example, the schema for the category of substance is said to be the capacity
to represent “the permanence of the real in time” (A143/B183). Since what is
permanent does not come to be or pass away, the schema makes it possible to
represent a substance — not merely as the subject of real attributes, but — as the
underlying element in temporal processes; and conversely, to represent what changes
in such processes as opposing states of a single thing, i.e. a substance.
3. There are three basic modal categories: possibility, actuality and necessity. At A144
f./B184, Kant gives their schemata as, respectively,
(P) The agreement [Zusammenstimmung] of the synthesis of differentrepresentations with the conditions of time in general. . . thus thedetermination of the representationof a thing at some time.
(A) Existence in [in] a determinate time.(N) Existence . . . at [zu] every time.15
13Ultimately, the categories will also have to be made to represent spatial existence (A220/B268).But this does not mean — as Paul Guyer has suggested — that “there is not a perfect match betweenKant’s initial definitions of the schemata for the categories [. . . ] and the claims about the conditionsfor the actual use of the[se] concepts [. . . ] that he will subsequently make” (2006), 100. The schematathat Kant introduces at A144/B184 pertain to the concept of existence in general; how they pertain toouter existence is a further question, requiring further reflection.14For an attempt at a reconstruction of such a deduction, see Allison (2004), 219 ff.15The Table of Categories also contains negative modalities: impossibility, non-being (or non-actuality)and contingency (see 2.5.2, above). Apparently their schemata can be formed by appropriate negationsof P, A and N. (For example, at A788/B816, Kant indicates that the schema for contingency is existenceat some, but not every, time.) He does not discuss these categories at A144 f./B184, however, and Iwill leave them aside here.
3. The Modal Schemata 51
Here existence is meant in the general sense I described in Chapter Two (i.e. that of
objective being); and the schemata for possibility, actuality and necessity are capacities
to use this concept differentially in relation to time.
The ultimate purpose of this chapter is to elucidate these capacities. Before I
can do this, though, I need to say something more about the pure schemata. That is
the purpose of the next two sections.
3.3
1. According to Kant, these schemata have two closely related functions. The first is
to “realize the categories”; the second is to “restrict them, i.e. limit them to
conditions that lie. . . in sensibility” (A146/B185 f., emphasis added).
I want to make some remarks about each function in turn; in section 3.4, I will
say a little about how they are related.
2. The schemata realize the categories by representing them in such a way that they
can be used in real judgments (where this implies, among other things, a
determination related to experience). For example, that the schema of actuality is
existence in a determinate time means that we can make real judgments on the actual
if we represent it that way.
I should emphasize that schematism is not itself real judgment (in the relevant
sense).16 One can of course say that “the schema of actuality is existence in a
determinate time”. And this is important in preparation for the Doctrine of Judgment.
But it is only preparation. To say that we can use a concept in relation to experience
if we represent it thus-and-so is not yet to use the concept in relation to experience.
3. It may seem strange that schematism, so conceived, should restrict the use of the16Beck appears to think otherwise (1967), 241 f. There is little textual support for his reading,however, and it is not required by the account.
3. The Modal Schemata 52
categories. But a concept can be restricted in two ways.
a. Determination. When a concept is determined in a synthetic judgment, it is
represented under a restriction corresponding to the predicate that is added to it.
For example, in the judgment “Every body is divisible”, the concept of a body is
restricted to divisibles. This does not mean that indivisible body is a contradictio in
adjecto (for then the judgment would be analytic).17 It means that there are no
indivisible bodies — they do not exist.18
b. Schematism. Because a schema does not add anything to the concept associated
with it, the restriction it imposes is only indirect. In principle, a concept might be
usable across various incommensurable conditions, so that it could be schematized
in various ways. As a matter of fact, Kant thinks there is only one way we can use
concepts — viz. in relation to time — and our actual schemata are exclusive. In
this sense, they can be said to restrict the use of the categories. But it would be
more accurate to say that they correspond to such a restriction.
3.4
1. What is the corresponding restriction?
Earlier I said that existence is temporal in relation to our knowledge; it is not,
however, temporal as such. Time is rather a formal representation of our sensibility
— Kant will say that it is the form of sensibility (A22/B36 et passim). And that is
ultimately the source of the restriction that Kant associates with the schemata. It is
a determination — not of existence as such, but — of our own mind.17Kant indicates that “Every body is divisible” is synthetic at B128. See also 2.3.4, above.18One can of course use the concept in negative judgments, e.g. that no indivisible body exists. But— following a long tradition — Kant holds that such judgments are not knowledge in themselves,even if they are true; their purpose is “merely to prevent error. Accordingly, negative propositionsthat would prevent a false judgment where no error is possible. . . are indeed empty — i.e. not at allsuited to their end — and for this reason very often ridiculous” (A709/B737; cf. A72/B97).
3. The Modal Schemata 53
Because the representation of time belongs to our subjective constitution,
however, it can be hard to see how a schema could realize the concept associated
with it; it might seem more natural to say that it idealizes the concept, since it
provides for the concept’s relation to our form of experience, our sensibility.
Evidently Kant does not want to say this; but why not?
2. I said that a concept is realized when it is represented in such a way that it can
be used in real judgments. But there are two factors in this representation. There
is, first, a sensible factor: the concept has to be schematized. And second, there is an
intellectual factor: the concept must be brought under the appropriate categories
(substance or accident, possibility or actuality, and so forth), since these are
representations of real objects as such. I want to say a little about this second factor;
then I will return to the first.
3. In Chapter Two, I argued that the categories are concepts of existence in the
general sense.19 In bringing a given concept (such as body) under a category (such
as substance) it is represented as the concept of an existing object (i.e. a thing in
itself). I also said that existence in this sense has a two-fold use: on one hand, an
object exists in relation to other objects; on the other, it exists in relation to the
subject, or at least his cognitive faculty. The representation of the former is made
possible by the categories of relation; the representation of the latter is made
possible by the categories of modality.
But I also argued that relation and modality are not simply two species of
existential representation; while relation is a representation of the determinate
content of judgments, modality is a representation of their mere form, and so is the
fundamental condition of all judgments, whether or not they determine their
content. That is why a judgment can have a modality even if it does not represent19See 2.3.5, above, for some qualifications.
3. The Modal Schemata 54
an objective relation — for example, “There is a body”.20
Since, however, modality is a merely subjective representation, it follows that
the fundamental intellectual condition of all judgments is ideal — even apart from
any consideration of the sensible condition under which the power of judgment can
be exercised.
4. The idealism that Kant associates with his philosophy is not merely a
consequence of his analysis of human sensibility, but a characterization of the act of
knowledge more generally: as originating not in the object but in the mind itself.21
(This is sometimes called the Copernican Hypothesis.22) This does not mean it is
unimportant that time is ideal; the point is rather that the ideality of time is only
one part of a broader idealism, an idealism that for Kant is inseparable from
philosophy itself.23
Schematism, then, is not idealization in the sense that it introduces a kind of
subjectivity into a concept; the subjectivity was already there, in the mere functions
of judgment. The schema simply represents the use of these functions in relation to
the sensibility of the very same subject.24
20See 2.4.2, above.21This is sometimes overlooked. Guyer, for example, takes Kant’s idealism to consist merely in theclaim that “a spatial and temporal view of things as they really are in themselves, independent of ourperceptions of them, would be demonstrably false” (1987), 333.22Cf. B n.:[T]he central laws of the movement of heavenly bodies — which Copernicus at first hadassumed only as a hypothesis, and which at the same time yielded certainty and proofof the invisible power (the Newtonian attraction) that binds together the structure of theworld — would have remained forever undiscovered if the latter had not ventured —in a manner contrary to the senses but nonetheless true — to seek the movements heobserved not in the objects of the heavens, but in the spectator.
The metaphysical analogue of Copernicus’s hypothesis is that “objects must conform to our knowl-edge” (B).23Cf. Natorp: “Philosophy, in its strictest historical sense, is not other than: idealism” (1903), .24It is true that Kant sometimes invites us to imagine forms of sensibility different from our own(e.g. A230/B283) — and this might suggest a more radical subjectivism than was originally envisaged.I cannot address this suggestion here, except to report Kant’s remark that the possibility of a non-temporal sensibility is one that “we cannot conceive [erdenken] or make comprehensible to ourselvesin any way” (A230/B283). The concept of possibility itself, he thinks, has real use only in relation totime (see 3.5, below). By contrast,
3. The Modal Schemata 55
5. It is this relation to sensibility that connects the two functions of the schema: it
simultaneously realizes the concept by making it usable in judgments under
sensible conditions, and restricts it in accordance with the form of sensibility itself
(A147/B187).
3.5
The schemata for the modal categories — P, A and N — can thus be understood as
relating these concepts to our own subject represented as a faculty of knowledge
under sensible conditions.
And this gives a richer sense to the progress of knowledge I have identified
with modal distinctions. It is not merely a progress from the possibility of knowing
to its actuality, and from this actuality to its necessity — a progress characteristic of
finite understanding in general. It is a progress in representing the relation of
existence to time, as the peculiar form of our sensibility. To be merely possible in
this sense is to bear a possible relation to time; to be actual is to bear an actual
relation to time; to be necessary is to bear a necessary relation to time.25
Conclusion
A reminder not to overstate the significance of this schematism. It is, as I said, only
preparation for the Doctrine of Judgment. I have characterized the sense that theabsolute possibility (that which is valid in every respect) is no mere concept of the under-standing, and can be of no use at all in experience, but belongs to reason alone, whichgoes beyond all possible use of the understanding. We have thus had to content ourselveswith [this] merely critical remark. . . [A232/B285.]
25My way of putting things differs from Kant’s. For him the schemata of the modal categoriescorrespond to a progress in the representation of the determinacy of existence in relation to time. Onecan certainly say this — determination has a number of senses. I have avoided it, however, becauseI want to avoid any suggestion that modal progress consists in introducing new representations (sc.determinations in the strict sense) into our knowledge. See 1.4.2, above.
3. The Modal Schemata 56
categories of modality must have in our judgments; I have not yet said what the
judgments are.
That is the purpose of the next chapter.
CHAPTER FOUR: THE PRINCIPLES OF MODALITY
[I]n behalf of the business which is in hand I entreat men to believe thatit is not an opinion to be held, but a work to be done.
—Bacon, The Great Instauration.1
4.1
1. In the previous chapter, I presented the schemata for the categories of modality,
which I described as preparation for the Transcendental Doctrine of Judgment. In
this chapter I am going to discuss the Doctrine itself, and in particular its
contribution to the theory of modality.
2. The purpose of the Transcendental Doctrine of Judgment is to explain how
experience is possible under the peculiar conditions of human sensibility. This is
accomplished, broadly speaking, through the assertion — and subsequent proof —
of various principles respecting the use of the pure schemata a priori. Although
these principles are formulated as general propositions (for example, that “In all
change of appearances, substance persists” [B224]), they should not be taken as
mere descriptions of facts. Their representation as principles of experience
constitutes its very possibility.2
The broad point will be familiar to readers of Kant, even if there is
disagreement on the details of its execution.3 Nonetheless, I think it has not been1Cited by Kant: B.2Of course, this is for the most part implicit; for the most part, the operation of our mind is hidden
from us. That is why philosophy can be interesting.3Among other things, there are questions about what experience should mean here; Van Cleve
(1999), 74 has distinguished eight possibilities.
4. The Principles of Modality 58
sufficiently appreciated — especially with regard to the principles of modality —
the Postulates of Empirical Thinking in General, as they are called (A161/B200).
They are the subject of this chapter.
3. At A218/B265 f., the Postulates are presented as follows:
(1) Whatever accords [übereinkommt] with the formal conditions ofexperience (with respect to intuition and concept) is possible.
(2) Whatever is connected [zusammenhängt] with the material conditionsof experience (sensation) is actual.
(3) That whose connection [Zusammenhang] with the actual isdetermined according to universal laws of experience is [ist] ([i.e.]exists [existiert] as) necessary.
Although he does not say it explicitly, Kant appears to intend for the modal
categories contained in these Postulates to be represented through their schemata.
So, for example, (1) means that whatever accords with the formal conditions of
experience is possible in the sense that it exists at some time. The Postulates are
thus principles of real rather than logical modality, and more specifically of
empirically (or phenomenally4) rather than absolutely real modality. In other words,
they are, like all the principles in the Doctrine of Judgment, propositions on human
experience.
4. But their constitutive role in experience has never been well understood; it is
widely thought that they do not have one.
This may have begun with Adickes’s confession that, “For me, the statement of
the Postulates is only explicable as an outgrowth of the architectonic [Anbequemen an
die Systematik]”5 Or perhaps it was Kemp Smith’s denunciation of
these so-called principles [as] not really principles [at all]; they merelyembody explanatory statements designed to render the preceding results
4See A146/B186: “The schema is properly only the phenomenon, or the sensible concept of anobject, in agreement with the category”.5(1887), 55.
4. The Principles of Modality 59
more definite. . . 6
In any case — and despite Paton’s pleas7 — this has become conventional wisdom.
Here is Bennett:
For the modal categories, [Kant] produces no ‘principles’ at all, but only‘explanations of the [modal categories] in their empirical employment’,giving a specious reason why this is all he can do.8
And Guyer:
Kant provides three ‘Postulates of Empirical Thought’. . . . These do notintroduce any new principles for empirical judgment at all but rather usethe principles already described to give sensible meaning to the logicalconcepts of modality, that is, possibility, actuality, and necessity.9
More recently, Gardner (1999), 128 has appeared to deny that the Postulates are so
much as synthetic judgments; Allison has denied it explicitly (2004), 225.
Longuenesse does not even discuss them (1998), (2005).
5. As a matter of interpretation, this is certainly uncharitable — Kant repeatedly
refers to the Postulates as principles,10 and he emphasizes their “far-reaching use
and influence” (A221/B268). It is not, however, completely misguided. The
Postulates are unlike the other principles Kant discusses in the Doctrine of
Judgment in at least two ways.
First, they are not accompanied by proofs. There is only a loosely arranged
“elucidation” (“Erläuterung”) of their transcendental significance, and of some of
their implications (A219 ff./B266 ff.) — for example, that “the grossness [of our
senses] has nothing to do with the form of possible experience in general”6(1918), 392. He took this as a symbol “of the perverting influence of Kant’s architectonic, as well
as of the insidious manner in which the older rationalism continued to pervert his thinking in his lesswatchful moments.”7(1936), 1.370 f.8(1966), 92.9(1987), 185.10Cf. A161/B200, A180/B223, A219/B266, A232/B285.
4. The Principles of Modality 60
(A226/B274). This suggests the Postulates are supposed to be self-evident in a way
the other principles are not. (The word postulate itself suggests this11 — though it
will turn out that Kant means something special by it.)
Second, he himself notes that the categories of modality “have this peculiarity”:
that they “do not in the least enlarge the concept to which they are attached as
predicate, but only express its relation to the faculty of knowledge” (A219/B266).12
In other words, “the principles [Grundsätze] of modality are not objectively
synthetic. . . but only subjectively” (A233 f./B286). To say that something, an object
or a determination, has a particular modality is just to say that the knowledge of it
comes under that modality — which is not, therefore, further knowledge of the
object.13 It is, rather, knowledge of one’s own mind — it is self -knowledge.
In fact it is a special kind of self-knowledge. For it is not just knowledge of an
object that happens to be oneself, or even knowledge that one is in a privileged
position to have just by being the self one is (for example, that one’s limbs are
situated in a certain way). It is knowledge that constitutes its object merely in the
knowing of it. I will call this self-constituting knowledge. And I want to suggest that
it is because the Postulates are self-constituting in this sense that Kant thinks they
cannot be proved.
4. I do not just mean that — as self-knowledge — the Postulates are (or are
supposed to be) self-evident. It is true that a proof is only needed (as Kant puts it)
to remove “the suspicion [that these principles are] merely concealed dogma[s]”
(A149/B188). If no such suspicion could arise, there would be no need for a proof;
everyone would just see that it was true.
But the Postulates are not self-evident — or, at least, not self-evidently so.14 It11“Postulate—a proposition. . . that one can require of another without proof” (Wolff, Vollständiges
mathematisches Lexicon, 1032). See, however, A232/B285.12This was discussed in 2.4 and 2.5, above.13For example, to say that Caius is actually a man will mean that the judgment “Caius is a man” isactual in the mind. It does not correspond to a property of Caius himself.14Here it may be useful to recall Lewis’s distinction between what is obvious and what is “obviously
4. The Principles of Modality 61
might be possible to bring oneself to the point of just seeing that (for example)
possibility consists in agreement with the formal conditions of experience; but this
will take some work. The question is why the work would not then be a proof.15
4.2
1. To answer this question, I want to say something about the Postulates themselves
— and in particular, their appeal to the formal and material conditions of
experience.
2. Kant associates the formal conditions of experience with the Postulate of
possibility, and the material conditions of experience with the Postulate of actuality.
He does not say that this is new or controversial. Indeed, he brings no attention to
it all. Nonetheless — it was new and controversial.
In the tradition, the concept of form was generally associated with actuality,
and the concept of matter with possibility. This goes back to Aristotle: “matter,” he
had said, “exists in a potential state, just because it may come to its form; and when
it actually exists, then it is in its form.”16
An example may make this clearer. Consider a bronze statue of a horse. Here
the matter will be the bronze; the form will be the horse-shape. Of course, this
shape is not an actual statue, for by itself it has no matter. Nonetheless, Aristotle
thinks, when we say that someone is making a horse statue — i.e. making it actual
— we mean that he is giving form to the bronze. We do not mean that he is givingobvious" (1990), 24.15Obviously the Postulates cannot be proved by the Leibnizian method of substituting identities(see “Primary Truths”). Because the propositions that Kant discusses in the Analytic of Principles aresynthetic, however, his proofs of them all appeal to propositions that can be denied without inducingabsurdity. It follows that — if self-evidence were supposed the only interesting difference betweenthe Postulates and the other principles — the difference between proof and elucidation could only be amatter of degree.16Metaphysics Θ.8, 1050a15 f.
4. The Principles of Modality 62
bronze to the shape.17
3. I do not think Kant would disagree with Aristotle about the statue. He is not
claiming that the concept of form can never be associated with actuality. He is
claiming, more narrowly, that — in the act of knowledge — the form is associated
with possibility, and the matter with actuality. And this, I take it, is connected with
his distinctive conception of knowledge.
In Chapter One, I argued that — according to Kant — the possibility of
knowledge is a (more or less general) capacity to know; the actuality of knowledge is
the act of such a capacity. If this is right, then the claim here is that the capacity is
associated with the form of knowledge, while the act is associated with the matter.
Now, this could be taken to mean that, before a capacity can be actualized, some
sort of material has to be added. Since Kant holds that the matter of knowledge is
sensation, it would follow that the capacity is actualized when, and only when, the
mind is in some way affected.18 Here is how Béatrice Longuenesse puts things:
Following Baumgarten, Kant writes that a conatus is associated withevery Vermögen [capacity or faculty].19 This conatus is a tendency or effortto actualize itself. For this tendency to be translated into action, it mustbe determined to do so by external conditions.20
Some passages in the Critique support this interpretation — for example, that
the faculty of knowledge is awakened to its exercise [Ausübung] byobjects that affect [rühren] our senses. . . [i.e.] the raw material of sensible
17Might someone mean that? Someone could of course say, “I have an idea for a statue, I just needsome bronze.” But the idea of a statue is not the same as its form, for the envisaged statue will haveboth form and matter. What is needed to realize it is not bronze, but — a statue.18See A19 f./B34: “The effect of an object on the capacity for representation, so far as we are affectedby it, is sensation.”19In fact, Kant did not write this. Longuenesse is citing a transcript of lectures from 1784-5known as the Metaphysics Volckmann ( 28.1: 434). Even if the transcript is accurate (and there isstrong reason to doubt this, for it is at best a second-hand copy of one student’s notes), it shouldbe be treated with care. Kant’s metaphysics course closely followed Baumgarten’s Metaphysica, andin many passages it is difficult to tell whether he is speaking for himself or simply explicating thetextbook for the benefit of his students. (By 1781, he had rejected nearly all its essential claims.) Kantmay have been following Baumgarten, as Longuenesse suggests; but he may also have been quotinghim; or the passage might simply have been invented by Volckmann or someone else.20(1998), 7
4. The Principles of Modality 63
impressions, [which it then] work[s] up into a knowledge of objects. . .[B1]
Presumably, then, there is a sense in which Longuenesse is right. But there is a
deeper sense in which she is wrong.
In the Appendix to the Transcendental Analytic — the Amphiboly of Concepts
of Reflection — Kant says that the concept of form “signifies. . . the determination”;
the concept of matter, by contrast, “signifies the determinable in general”
(A266/B322). In the example of the statue, the bronze would be the determinable,
while the horse-shape would be its determination. To the extent that external
conditions determine the capacity to act, then, they will be the formal element in
knowledge; the capacity that they determine will be the material element. This is a
possible theory of knowledge, but it is not Kant’s theory.21
Now it could be that Longuenesse is using the word determine loosely, and what
she really means is that external conditions have to be in place for the capacity for
knowledge to be actualized.22 That is probably true in some sense; it certainly
agrees with B1, and also with Kant’s more general claim that, if the right sort of
external conditions were not in place, the understanding “would not happen
[stattfinden]”23 (A654/B682; cf. A100 f.). But it does not capture the meaning of the
Postulates.
If a capacity for knowledge could not be actualized except under certain
external conditions, then by itself it would not be possible knowledge. Rather, the
possibility of knowledge would lie in the external conditions (the matter) — which
the capacity would work up into actuality. But this is not what the Postulates say.21I do not think it is Baumgarten’s theory either, though this is less clear.22The reason she uses the word determine is that Kant (reportedly) used it in the Volckmann lectures:“the conatus [or] striving,” he is reported to have said, “is properly speaking the determination of acapacity ad actum” ( 28.1: 434). I am not sure what this means. But one thing it does not mean isthat external conditions determine the capacity. The determination is supposed to be the “conatus” —and whatever that is, it is not an external condition.23The expression is no less strange in German: the difficulty lies not in the language, but in thethought — namely, that the understanding is not a thing, but a kind of event.
4. The Principles of Modality 64
What they say is that the possibility of knowledge lies in its mere form; the
actuality lies in the matter. This is what has to be explained, and what Longuenesse
does not properly explain.
The proper explanation is that the capacity for knowledge determines — or
constitutes — itself. It makes itself actual. I discussed this Chapter One, but it is
worth mentioning again. A capacity for knowledge is typically associated with a
problematic judgment. This judgment represents a particular thought — for
example “Every body is divisible” — as merely possible. The act of this capacity is
an assertoric judgment — a judgment that represents the thought as actual. Now, in
Chapter One I argued that this act always occurs spontaneously. This does not
mean that it always occurs immediately, or without reference to experience. Human
understanding is imperfect; there is a sense in which understanding takes time.24
But it does mean that, once a capacity for knowledge has been acquired, no
exogenous material has to be added for it to act. In other words, because the
capacity is self-constituting, it already contains the form of knowledge — for otherwise,
it would have to be constituted by something outside it (viz. by external
conditions). Since, however, it is in the first instance a mere capacity, the knowledge
is, at least initially, only possible.
This explains how a capacity for knowledge can be associated with both form
and possibility. But why is the actualization of this capacity associated with the
matter? Why is the matter not already in the capacity?
In a way, it is. As I said in Chapter One, a capacity for knowledge can only be
acquired under certain external conditions. That is why the Postulates all appeal to
experience; and it is also why the remarks of B1 and A654/B682 are relevant here.
The understanding, which is the capacity for knowledge in general, would be
dormant in the absence of any sensation.25 It would not (in his word) happen. But24It would probably be more accurate to say that coming to understand takes time.25At A654/B682, Kant adds that the understanding would be dormant even in the presence of
4. The Principles of Modality 65
— for just this reason — the “awakening to exercise” that he describes is more
properly described as the acquisition of a capacity (viz. the understanding as a
faculty of judgment). This is very different from its act. External conditions only
bring the understanding into a position where it can know things — they enable it
to be a capacity for knowledge at all. That is the sense in which the matter is
already in the capacity.
But since the capacity thus acquired is (at least initially) only a capacity, and
not yet an act, the awareness of its material conditions is (at least initially) only a
possible awareness. In other words, knowledge is initially represented as merely
possible with respect to its matter. It is knowledge of merely possible existence. It is
only when the capacity for knowledge acts that the matter will be represented as
actual. Indeed, the act of the capacity just is this representation.
Accordingly, when Kant says that “whatever is connected with the material
conditions of experience is actual” he does not mean that these conditions make the
knowledge of an object actual, or even that they are conditions of its act. He simply
means that actual knowledge is knowledge of these conditions — as actual.
3. This reading of the Postulates implies that the difference between the knowledge
that something is possible and the knowledge that it is actual is entirely subjective
— that it lies entirely within the mind. This does not mean that whether something
is possible or actual is a matter of personal preference. More generally, it is not a
psychological difference, a difference in the way things appear relative to a particular
mental state.26 Rather, it is a difference in the character of the knowledge itself —
the way it figures in subjects in general. In the next section, I will consider some
implications of this point.sensations if they were too varied in their content, or not varied enough. For him, the possibility ofhuman knowledge is fundamentally contingent.26See 1.3, above, for a discussion.
4. The Principles of Modality 66
4.3
1. So far I have characterized capacities for knowledge as problematic judgments,
and their acts as assertoric judgments. But there is also the capacity for knowledge,
and its act. The former is the understanding as the faculty of judgment (A69/B94);
the latter is the power of judgment.27 (The capacity to perfect this power is called
reason.28)
And — as Kant will eventually suggest — these general capacities, or faculties,
are the real object of the three Postulates. For they represent, respectively,
how [the object] (together with all its determinations) is related [sich. . .verhalte] to the understanding and its empirical use, to the empiricalpower of judgment and to reason (in its application to experience).[A219/B266.]29
I will try to explain this.
2. There are two ways particular judgments can be related to a faculty of
knowledge.30
a. The faculty can be derived from the particular judgments. This is the case with what
could be called the empirical faculties of knowledge — general capacities that can be
acquired in experience. For example, someone who comes to know a lot about
geography — the capitals of the various countries, their major cities, rivers,
topographies and so forth — could be said to have acquired a geographical faculty.27See 3.1.1, above.28To avoid complications, I will mainly avoid discussing reason here. I simply want to note thatKant includes it in this context.29Cf. A75 n./B100 n.: “It is just as if thinking [in a problematic judgment] were a function of theunderstanding, [in an assertoric judgment] a function of the power of judgment, [and in an apodeicticjudgment] a function of reason.”30Cf. B124 f:There are only two possible cases in which a synthetic representation and its object cancome together, relate to one another in a necessary way, and as it were meet up: either ifthe object alone makes the representation possible, or if the representation alone makesthe object possible.
4. The Principles of Modality 67
(This is normally expressed by saying “he’s good at geography”.) And this may be
useful to know about him if, for example, there is a shortage of teachers of
geography.
But it will not be useful as an explanation of his geographical knowledge. For
example, it will not be useful to explain his knowledge of Baltic port cities by
appealing to a geographical faculty; the reverse, in fact. He is said to have a
geographical faculty only because he knows things like: the location of port cities
along the Baltic. This is true of empirical faculties of knowledge in general. They
have no explanatory role beyond the particular judgments associated with them.
(They are nothing but particular judgments, united under a common theme.)
b. The particular judgment can be derived from the faculty. In this case, the faculty will
be pure; it will belong to the mind a priori. And the understanding is a faculty of
this kind. Although it always takes a particular shape in a particular subject —
there can, after all, be disagreements about how to understand something —
understanding as such is not an aggregate of particular judgments. It is the faculty of
judging; it makes judgments possible. If it were the other way around, there could
not be a priori knowledge; every judgment would depend on sensible material.31
3. This brings me to the purpose of the Postulates. In connecting the faculties of
knowledge with existence, they do not just explain the meaning of the modal
categories under sensible conditions. They make the representation of existence
possible in the first place.
Here is how Kant puts it:
If [for example] I represent to myself a thing that is permanent, so thateverything in it that changes merely belongs to its state [i.e. a substance],I can never know from such a concept alone that such a thing is possible.
31Although it is sometimes said that Kant wants to show that a certain kind of a priori knowledgeis possible in the Critique, in fact he takes its possibility for granted (B3). What he wants to show ishow this knowledge is possible (B19).
4. The Principles of Modality 68
. . . It is only because [this] concept[. . . ] expresses the relations ofperceptions in every experience a priori that we know its objective reality[i.e. its empirical possibility]. . . [A221/B268 f.]
Commentators have tended to focus on the appeal to experience here, and the
attendant distinction of logical from real possibility. And that is certainly important:
the possibility of a concept is not the same as the possibility of something’s existing
under that concept. But there is a deeper point, too.
The relation of a given concept to experience is not, in general, something that
can be discovered by examination of the concept alone. In some cases, of course, it
is possible. For example, it is obvious that there cannot be non-simultaneous
coexistence because this cannot be constructed in time.32 But constructibility in
time is not a sufficient criterion of real possibility; many concepts can be constructed
in time that are nonetheless really impossible.33
The real possibility of a concept can only be known from its relation to one’s
own faculties of knowledge. Whether a concept belongs to possible experience can
only be known from its relation to the understanding; whether it also belongs to
actual experience can only be known from its relation to the power of judgment.
And the role of the Postulates is to effect these relations. I have in mind something
like the following:
(1’) Whatever accords with the formal conditions of experience is, invirtue of this very principle, possible.
(2’) Whatever is connected with the material conditions of experience is,in virtue of this very principle, actual.
(3’) That whose connection with the actual is determined according touniversal laws of experience is, in virtue of this very principle,necessary.
This is just to say that the possibility of knowledge is effected by the awareness of
the relation in virtue of which it is possible; the actuality of knowledge is effected32Cf. B256 f.: “Things are coexistent [zugleich] if . . . the perception of one can follow the otherreciprocally.”33Kant gives some examples at A222/B270.
4. The Principles of Modality 69
by the awareness of the relation in virtue of which it is actual. The Postulates, as I
understand them, express this awareness in its most general shape.
4.4
1. At the beginning of this chapter, I asked why Kant thinks the Postulates cannot
be proved, but only elucidated. Now I want to address this question.
The answer is that they are quasi-practical principles. Obviously they are not
practical in the sense that they relate to human action, or fall under principles of
practical reason. But they are quasi-practical in the sense that they bring into
existence the very concepts that they represent.34 Kant puts it like this: “The
principles of modality. . . predicate [sagen] nothing of a concept other than the act
[or action: Handlung] of the faculty of knowledge through which [the very same
concept] is generated” (A234/B287).35
A proposition of this kind, he says, is called a postulate. This does not
necessarily mean that it is “immediately certain” (A233/B285).36 A postulate, as
Kant understands it, is a procedure for generating a certain concept. He gives an
example from Euclid: “with a given line, to describe a circle on a plane from a
given point”.37 The reason this postulate cannot be proved is not that it is
self-evident, but that the concept of a circle — which any proof would presuppose
— was generated by the postulate itself.
Now, Kant says,
[W]e can, with just the same right, postulate the principles of modality,since they do not enlarge the concept of things in general, but only showthe manner in which [this concept] in general is connected with thepower of knowledge [Erkenntniskraft]. [A234/B287].
34Kant himself uses the word “practical” to describe postulates in general at A234/B287.35This point bears comparison with Kant’s discussion of the fact of reason of the second Critique( 05: 55 f.).36See 4.1.5, above.37This is from the Elements, Postulate 3.
4. The Principles of Modality 70
In other words, the principles of modality cannot be proven — not because their
content cannot be doubted — but because they themselves generate this content.
They make possible knowledge possible, and actual knowledge actual. In this sense,
our knowledge is essentially self-constituting: it makes itself possible; it makes itself
actual. It is, in other words, essentially spontaneous.
Conclusion
The spontaneity of knowledge is the broadest and most important theme of the
Critique of Pure Reason. It is the primary characteristic of the understanding
(A51/B75); it is the ground of the possibility the Copernican Revolution, and the
resulting canon of knowledge a priori in which the Critical science consists.38
Human knowledge, however, is spontaneous in a limited sense;39 as I tried to
bring out in Chapter Two, all our knowledge depends on, and makes essential
reference to, experience. (This eventually proved to involve the representation in
time in a way I tried to bring out in Chapter Three.) And it is this idea of a limited
spontaneity that I have tried to connect with Kant’s theory of modality — for it is
through the modal function that our knowledge constitutes itself under empirical
conditions.
The theory is, of course, only one part of a much broader theory of existence: I
have said almost nothing about substances and their relations to one another, and
nothing at all about the practical knowledge of existence. There is more work to be
done; but there is always more work. That is the point.
Does the road wind up-hill all the way?Yes, to the very end.
38See A877/B849; Prol. sec. 1, 04: 265 ff. The scientific aspirations of the Critique are nicelycaptured by Merritt (2004).39I owe this description to Engstrom’s seminal (2006).
71
APPENDIX: THE FUNCTIONS OF MODALITY AND RELATION
.
Logical Functions Real functions
Modality: Problematic Possibility – Impossibility
Assertoric Existence – Non-being =Actuality – Non-actuality
Apodeictic Necessity – Contingency. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Relation: Categorical Substance and accident
Hypothetical Cause and effect
Disjunctive Reciprocity of agent & patient
72
BIBLIOGRAPHY1
Thine eyes did see my substance, yet being unperfect; and in thy book allmy members were written. . .
—Ps. 139.
Adickes, Erich (1887). Kants Systematik als systembildener Factor. Berlin: Mayer E.Müller.
Allison, Henry (2004). Kant’s Transcendental Idealism: an Interpretation and Defense(2nd edition). New Haven: Yale University Press.
Anderson, R. Lanier (2008). “Comments on Wayne Martin, Theories of Judgment”.Philosophical Studies 137.
Bäck, Allan (1987). “Avicenna on Existence”. Journal of the History of Philosophy 25,351-367.
Beck, Lewis White (1967). “Can Kant’s Synthetic Judgments Be Made Analytic?”In: Gram, Moltke S. (ed.). Kant: Disputed Questions. Chicago: QuadrangleBooks.
Bennett, Jonathan (1966). Kant’s Analytic. New York: Cambridge University Press.
Boyle, Matthew (). “Kant on Logic and the Laws of Understanding”.
Burnyeat, .. (2008). “Kinêsis vs. Energeia: A Much-Read Passage in (but Not of)Aristotle’s Metaphysics”. In: Sedley, David (ed.). Oxford Studies in AncientPhilosophy 34, 219-92.
Buroker, Jill Vance (2006). Kant’s Critique of Pure Reason: An Introduction. NewYork: Cambridge University Press.
Dummett, Michael (1981). Frege: Philosophy of Language. Cambridge, : HarvardUniversity Press.
Engstrom, Stephen (2006). “Understanding and Sensibility”. Inquiry 49, 2-25.
Forgie, J. William (2008). “How is the question ’Is Existence a Predicate?’ relevantto the ontological argument?” International Journal for Philosophy of Religion 64,117-133.
1Primary sources are given in the Preface, above.
Bibliography 73
Frede, Michael (1994). “Aristotle’s Notion of Potentiality in Metaphysics Θ”. In:Scaltsas, Theodore, Charles, David and Gill, Mary Louise (eds.). Unity, identity,and explanation in Aristotle’s metaphysics. New York: Oxford University Press.
Gardner, Sebastian (1999). Kant and the Critique of Pure Reason. London:Routledge.
Gibbons, Sarah (1994). Kant’s Theory of Imagination: Bridging Gaps in Judgement andExperience. New York: Clarendon Press.
Guyer, Paul (1987). Kant and the Claims of Knowledge. New York: CambridgeUniversity Press.
——— (1992). “The Transcendental Deduction of the Categories”. In: Guyer (ed.).The Cambridge Companion to Kant. New York: Cambridge University Press.
——— (2006). Kant. New York: Routledge.
Grier, Michelle (2001). Kant’s Doctrine of Transcendental Illusion. New York:Cambridge University Press.
Jachmann, .. (1804). Immanuel Kant geschildert in Briefen an einen Freund.Königsberg: Friedrich Nicolovius.
Kahn, Charles (1972). “On the Terminology for Copula and Existence”. In: Stern,.., Hourani, Albert Habib and Brown, Vivian (ed). Islamic Philosophy and theClassical Tradition: Essays Presented by His Friends and Pupils to Richard Walzeron His Seventieth Birthday. Oxford: Oxford University Press.
——— (1973). The Verb ‘Be’ in Ancient Greek. Dordrecht: Reidel.
Kemp Smith, Norman (1918). A Commentary to Kant’s “Critique of Pure Reason”.London: Macmillan.
Kneale, William and Kneale, Martha (1985). The Development of Logic. New York:Oxford University Press.
Lewis, David (1990). “Noneism or Allism?”. Mind 99, pp. 23-31.
Kosman, .. (1994). “The Activity of Being in Aristotle’s Metaphysics”. In: Scaltsaset al. (op. cit.).
Longuenesse, Béatrice (1998). Kant and the Capacity to Judge: Sensibility andDiscursivity in the Transcendental Analytic of the “Critique of Pure Reason".Princeton: Princeton University Press.
——— (2005). Kant on the Human Standpoint. Cambridge: Cambridge UniversityPress.
Bibliography 74
Maier, Anneliese (1930). Kants Qualitätskategorien. Berlin: Kant-StudienErgänzungshefte 65.
Malter, Rudolf (ed.), (1990). Immanuel Kant in Rede und Gespräch. Hamburg: FelixMeiner.
Martin, Wayne M. (2006). Theories of Judgment: Psychology, Logic, Phenomenology.New York: Cambridge University Press.
Merritt, Melissa McBay (2004). Drawing from the Sources of Reason: ReflectiveSelf-Knowledge in Kant’s First Critique. Doctoral dissertation, University ofPittsburgh.
Natorp, Paul (1903). Platos Ideenlehre: eine Einführung in den Idealismus. Leipzig:Felix Meiner.
Owen, David (2007). “Locke on Judgment”. In: Newman, Lex (ed.). The CambridgeCompanion to Locke’s Essay. New York: Cambridge University Press.
Paton, .. (1936). Kant’s Metaphysic of Experience (2 vols.). London: George Allenand Unwin.
Reich, Klaus (1935). Kant und die Ethik der Griechen. Tübingen: Mohr Siebeck.
Rescher, Nicholas (1964). “Aristotle’s Theory of Modal Syllogisms and ItsInterpretation”. In: Bunge, Mario (ed.). The Critical Approach to Science andPhilosophy: Essays in Honor of Karl Popper. New York: Free Press.
Rödl, Sebastian (2005). Kategorien des Zeitlichen: eine Untersuchung der Formen desendlichen Verstandes. Frankfurt am Main: Suhrkamp.
Rosenkoetter, Timothy (2009). “Absolute Positing, the Frege Anticipation Thesis,and Kant’s Definitions of Judgment”. European Journal of Philosophy 18,539-566.
Sellars, Wilfrid (1968). Science and Metaphysics: Variations on Kantian Themes.London: Routledge and Kegan Paul.
Sutherland, Daniel (2004). “The Role of Magnitudes in Kant’s Critical Philosophy”.Canadian Journal of Philosophy 34, 411-441.
——— (2005). “The Point of Kant’s Axioms of Intuition”. Pacific PhilosophicalQuarterly 86, 135-159.
Tonelli, Giorgio (1966). “Die Voraussetzungen zur Kantischen Urteilstafel in derLogik des 18. Jahrhunderts”. In: Kaulbach, Friedrich and Ritter, Joachim(eds.). Kritik und Metaphysik, Heinz Heimsoeth zum achtzigsten Geburtstag.Berlin: Walter de Gruyter.
Bibliography 75
Van Cleve, James (1999). Problems from Kant. New York: Oxford University Press.
Vuillemin, Jules (1960-1). “Reflexionen über Kants Logik”. Kant-Studien 52,310-355.
Wood, Allen (1978). Kant’s Rational Theology. Ithaca: Cornell University Press.
Zöller, Günter (2008). Kant and the problem of existential judgment: critical commentson Wayne Martin’s Theories of Judgment. Philosophical Studies 137, 121-134.